1 00:00:00,320 --> 00:00:03,760 Speaker 1: The Big Bets on Campus Podcast Podcast. 2 00:00:04,160 --> 00:00:12,760 Speaker 2: Pot all right, here we go. Fifteen Cundy, twenty five, 3 00:00:12,800 --> 00:00:15,440 Speaker 2: thirty thirty five Horny Cody five fifty. 4 00:00:15,280 --> 00:00:20,440 Speaker 3: Hand the kick as block the college football ware stunner. 5 00:00:23,200 --> 00:00:29,080 Speaker 3: Oh my goodness, he stumbled and bumble the lateral turn 6 00:00:29,080 --> 00:00:33,839 Speaker 3: the counter of the job, do you the lad I'm not. 7 00:00:34,040 --> 00:00:40,199 Speaker 4: Gonna table off the field to night. It's not the 8 00:00:40,400 --> 00:00:42,479 Speaker 4: size of the dog in the fighting. 9 00:00:42,760 --> 00:00:45,080 Speaker 3: It's the size of the fight in the dog. 10 00:00:48,640 --> 00:00:51,239 Speaker 2: Welcome to the Big Bets on Campus Podcast. This is 11 00:00:51,240 --> 00:00:53,920 Speaker 2: a special episode. We're squeezing in right under the wire 12 00:00:54,160 --> 00:00:57,040 Speaker 2: before the season starts in Earnest on Thursday night. I'm 13 00:00:57,080 --> 00:00:59,920 Speaker 2: my Calvary's. I'm joined today by Kelly Ford, the care 14 00:01:00,640 --> 00:01:03,520 Speaker 2: of the k Forward ratings. These ratings sought every team 15 00:01:03,600 --> 00:01:06,160 Speaker 2: number one Texas all the way down to lowly Kent State, 16 00:01:06,360 --> 00:01:08,399 Speaker 2: bringing up the rear at the one hundred and thirty 17 00:01:08,480 --> 00:01:12,319 Speaker 2: six rated FBS team. Kelly, Welcome to the show. Brother. 18 00:01:12,400 --> 00:01:14,720 Speaker 2: College football is here and who better to discuss it 19 00:01:14,760 --> 00:01:15,080 Speaker 2: than you. 20 00:01:15,600 --> 00:01:17,000 Speaker 4: Hey, Mike, thanks for having me. 21 00:01:17,440 --> 00:01:19,600 Speaker 1: It's been a while since we've talked, so I'm excited 22 00:01:19,640 --> 00:01:22,840 Speaker 1: to be back with you today talking college football and yeah, 23 00:01:22,880 --> 00:01:23,520 Speaker 1: the season's here. 24 00:01:23,560 --> 00:01:25,200 Speaker 4: I mean Week zero is awesome. 25 00:01:25,240 --> 00:01:26,920 Speaker 1: It gives us that kind of taste we've been off 26 00:01:26,959 --> 00:01:29,600 Speaker 1: for months and months, so it's exactly what we need 27 00:01:29,640 --> 00:01:32,880 Speaker 1: on that first Saturday, but a full slate and it 28 00:01:32,920 --> 00:01:36,920 Speaker 1: bleeds into you know, Thursday, Friday, Sunday, Monday. This is 29 00:01:36,959 --> 00:01:39,600 Speaker 1: what college football fans live for. It's what I live for, 30 00:01:39,760 --> 00:01:42,200 Speaker 1: the start of a new college football season. Everybody has 31 00:01:42,240 --> 00:01:43,280 Speaker 1: their dreams in front of them. 32 00:01:43,400 --> 00:01:44,479 Speaker 4: I can't wait to get into it. 33 00:01:45,040 --> 00:01:46,600 Speaker 2: Before we get into some of the teams that you'd 34 00:01:46,640 --> 00:01:49,040 Speaker 2: like think are overrated, underrated on the season, even some 35 00:01:49,080 --> 00:01:51,320 Speaker 2: week one plays that your model is showing value in 36 00:01:51,560 --> 00:01:54,320 Speaker 2: Let's start with a two parter, what is your background 37 00:01:54,360 --> 00:01:57,560 Speaker 2: in sports analytics in particular? And the second part, how 38 00:01:57,600 --> 00:02:00,080 Speaker 2: does someone build something like you have from scratch? Are 39 00:02:00,120 --> 00:02:02,520 Speaker 2: you hearing voices in the cornfield like Kevin Costner and 40 00:02:02,520 --> 00:02:04,680 Speaker 2: field of dreams like where does this all come from? 41 00:02:04,720 --> 00:02:07,840 Speaker 2: This desire to try to quantify a sport that is 42 00:02:07,920 --> 00:02:10,959 Speaker 2: so messy, that can be so difficult to predict. Walk 43 00:02:11,000 --> 00:02:13,119 Speaker 2: me through the whole process from day one? 44 00:02:13,600 --> 00:02:16,079 Speaker 1: Yeah, I wish I was able to say I could 45 00:02:16,080 --> 00:02:18,160 Speaker 1: hear those voices. They could probably make a good movie 46 00:02:18,160 --> 00:02:20,320 Speaker 1: out of it, but my story is probably not quite 47 00:02:20,360 --> 00:02:21,160 Speaker 1: as interesting. 48 00:02:21,240 --> 00:02:23,799 Speaker 4: But you know my background and how I got into this. 49 00:02:24,360 --> 00:02:27,240 Speaker 1: College football has always been my favorite sport for as 50 00:02:27,280 --> 00:02:30,000 Speaker 1: long as I can remember. And I'm from Indiana, so 51 00:02:30,240 --> 00:02:31,960 Speaker 1: you know, in Indiana there's the saying it's like in 52 00:02:31,960 --> 00:02:34,680 Speaker 1: forty nine states, it's just basketball, but this is Indiana, 53 00:02:34,720 --> 00:02:37,200 Speaker 1: and I like basketball, but for whatever reason, from a 54 00:02:37,320 --> 00:02:38,959 Speaker 1: very very young age, college. 55 00:02:38,760 --> 00:02:40,440 Speaker 4: Football has been what I have loved. 56 00:02:40,480 --> 00:02:44,480 Speaker 1: In the two thousand and two Ohio State teams, that's 57 00:02:44,520 --> 00:02:47,360 Speaker 1: my team. My mom's family's from the state of Ohio. 58 00:02:47,400 --> 00:02:49,600 Speaker 1: They kind of ingrained in me before I could even walk, 59 00:02:49,960 --> 00:02:52,480 Speaker 1: is what the story goes. So when that twenty and 60 00:02:52,480 --> 00:02:54,880 Speaker 1: two team won the national championship, I was like, Okay, 61 00:02:55,120 --> 00:02:56,640 Speaker 1: this is it for me. And then the two thousand 62 00:02:56,680 --> 00:02:58,919 Speaker 1: and seven season, which will have State actually lost the 63 00:02:59,000 --> 00:03:03,120 Speaker 1: national championship, but that season was so incredible, so much chaos, 64 00:03:03,120 --> 00:03:05,640 Speaker 1: so many upsets. I was like, all right, this is 65 00:03:05,800 --> 00:03:08,440 Speaker 1: the sport. Nothing can top it for me. So I've 66 00:03:08,480 --> 00:03:12,640 Speaker 1: always had the love for college football, the analytics. My 67 00:03:12,760 --> 00:03:16,000 Speaker 1: undergrad degree is a mechanical engineering, so I've always kind 68 00:03:16,040 --> 00:03:20,640 Speaker 1: of had a knack for math. Statistics, logic, analytics like 69 00:03:20,919 --> 00:03:23,320 Speaker 1: that all kind of comes naturally to me. And so 70 00:03:23,600 --> 00:03:26,040 Speaker 1: I've been a big fan of Bill Connolly and Brian 71 00:03:26,080 --> 00:03:29,400 Speaker 1: FROMO for a really long time. And I found myself, 72 00:03:29,440 --> 00:03:32,000 Speaker 1: you know, in high school in college talking with my 73 00:03:32,080 --> 00:03:35,160 Speaker 1: friends about college football, kind of through the Bill and 74 00:03:35,440 --> 00:03:37,640 Speaker 1: the Brian lens of S at the time, S and 75 00:03:37,680 --> 00:03:42,280 Speaker 1: P plus FEI, and I thought, you know, I follow 76 00:03:42,360 --> 00:03:45,280 Speaker 1: these guys, I keep up with their work. I'm as 77 00:03:45,280 --> 00:03:48,600 Speaker 1: into it as anybody. I think I could give this 78 00:03:48,680 --> 00:03:50,720 Speaker 1: a shot and see what I can come up with. 79 00:03:50,840 --> 00:03:52,920 Speaker 1: Instead of just referencing Bill and Brian all the time, 80 00:03:53,000 --> 00:03:54,640 Speaker 1: what if I came up with my own metrics. So 81 00:03:54,840 --> 00:03:57,720 Speaker 1: that's really how it started, was trying to just emulate 82 00:03:57,840 --> 00:04:00,240 Speaker 1: what those guys were doing. And as you build your 83 00:04:00,240 --> 00:04:02,920 Speaker 1: own data set, as you get more familiar with the information, 84 00:04:03,280 --> 00:04:05,000 Speaker 1: you kind of find your own way and you find 85 00:04:05,000 --> 00:04:07,400 Speaker 1: your secret sauces. And that's really what I've been able 86 00:04:07,440 --> 00:04:10,520 Speaker 1: to do in the last now geez, six or seven seasons, 87 00:04:10,840 --> 00:04:11,480 Speaker 1: and it's been fun. 88 00:04:11,520 --> 00:04:11,640 Speaker 4: You know. 89 00:04:11,680 --> 00:04:15,120 Speaker 1: I have the website Kfoardratings dot com. I'm on x 90 00:04:15,160 --> 00:04:17,360 Speaker 1: at kfward Ratings, and that's really where I post all 91 00:04:17,360 --> 00:04:20,160 Speaker 1: my things. So it's been it's been fun to see 92 00:04:20,160 --> 00:04:23,000 Speaker 1: it grow and to see the following kind of latch 93 00:04:23,080 --> 00:04:25,440 Speaker 1: onto it. And it's not all positive, and then I 94 00:04:25,480 --> 00:04:27,320 Speaker 1: get it that's part of social media and being on 95 00:04:27,360 --> 00:04:30,159 Speaker 1: the internet. But I really enjoy the interactions, you know, 96 00:04:30,200 --> 00:04:32,320 Speaker 1: more times than not. So that's how it got started. 97 00:04:32,320 --> 00:04:35,200 Speaker 1: It was really the convergence point of my skill set 98 00:04:35,320 --> 00:04:38,120 Speaker 1: of you know, math and data and science and then 99 00:04:38,320 --> 00:04:39,719 Speaker 1: my interest in college football. 100 00:04:40,520 --> 00:04:42,599 Speaker 2: Now you just mentioned your site, and for my money, 101 00:04:42,600 --> 00:04:44,520 Speaker 2: the coolest part about your site is the way that 102 00:04:44,560 --> 00:04:46,480 Speaker 2: you slice and dice the data. It's not just a 103 00:04:46,600 --> 00:04:49,200 Speaker 2: pure here's the ratings one to one hundred and thirty six. 104 00:04:49,440 --> 00:04:51,440 Speaker 2: You have all these different ways of approaching in one 105 00:04:52,080 --> 00:04:55,080 Speaker 2: particular data project that you know has grown a momentum 106 00:04:55,160 --> 00:04:58,320 Speaker 2: and also popularity over the last few seasons is your 107 00:04:58,360 --> 00:05:00,920 Speaker 2: watchability score. And this is I mean, that's really useful, 108 00:05:00,960 --> 00:05:03,839 Speaker 2: particularly for casual fans who are looking to say, hey, 109 00:05:03,920 --> 00:05:06,719 Speaker 2: I can't just watch seventeen hours of college football every 110 00:05:06,760 --> 00:05:09,760 Speaker 2: single Saturday. I need to pick my spots. And essentially 111 00:05:09,760 --> 00:05:11,520 Speaker 2: what you've done is you've created a score zero to 112 00:05:11,520 --> 00:05:14,120 Speaker 2: one hundred and it blends the overall quality of the teams, 113 00:05:14,400 --> 00:05:17,520 Speaker 2: the stakes involved in the likelihood of a competitive game. 114 00:05:17,560 --> 00:05:19,280 Speaker 2: Did they get the main components of that right? 115 00:05:19,839 --> 00:05:23,720 Speaker 1: Yeah, that's exactly right, Mike. And to the larger point, 116 00:05:24,240 --> 00:05:27,280 Speaker 1: just posting lists and posting texts. That's how I kind 117 00:05:27,279 --> 00:05:29,320 Speaker 1: of started out when I joined Twitter at the time 118 00:05:29,480 --> 00:05:31,560 Speaker 1: in twenty nineteen, I had no social media prior to that. 119 00:05:31,800 --> 00:05:32,640 Speaker 4: It's just not my thing. 120 00:05:32,880 --> 00:05:34,240 Speaker 1: But my friends were like, dude, you got to get 121 00:05:34,240 --> 00:05:35,560 Speaker 1: this out there. We're having a lot of fun, we're 122 00:05:35,560 --> 00:05:37,200 Speaker 1: making money using your numbers. I was like, all right, 123 00:05:37,200 --> 00:05:39,679 Speaker 1: So we tried it and I posted texts and lists 124 00:05:39,720 --> 00:05:41,920 Speaker 1: and all that and it was fine. But I kind 125 00:05:41,920 --> 00:05:44,880 Speaker 1: of learned people are drawn towards visuals. They want to 126 00:05:44,920 --> 00:05:47,600 Speaker 1: see a graphic. They want to see a snapshot of 127 00:05:47,680 --> 00:05:49,760 Speaker 1: their team and what does it mean? And they got 128 00:05:49,800 --> 00:05:51,359 Speaker 1: and you got to catch their attention and you know, 129 00:05:51,400 --> 00:05:53,279 Speaker 1: three to five seconds or their scroll and pass, and 130 00:05:53,320 --> 00:05:56,440 Speaker 1: so the watchability graphic has been something that people have 131 00:05:56,520 --> 00:05:57,760 Speaker 1: really gravitated towards. 132 00:05:57,880 --> 00:05:58,360 Speaker 4: And you're right. 133 00:05:58,600 --> 00:06:01,479 Speaker 1: Every single Friday I post it on X I posted 134 00:06:01,560 --> 00:06:04,080 Speaker 1: on the website and it is exactly what you said. 135 00:06:04,480 --> 00:06:08,560 Speaker 1: It's two parts combined into one. One overall score and 136 00:06:08,600 --> 00:06:10,960 Speaker 1: I've scaled it a different scales over the years. For 137 00:06:10,960 --> 00:06:12,560 Speaker 1: the last couple of years it's been a zero to 138 00:06:12,640 --> 00:06:15,440 Speaker 1: ten scale. Ten point zero out of ten is like 139 00:06:15,600 --> 00:06:16,560 Speaker 1: maximum watchability. 140 00:06:16,600 --> 00:06:17,240 Speaker 4: You can't miss it. 141 00:06:17,480 --> 00:06:21,080 Speaker 1: Spoiler alert Texas at Ohio State this weekend. Ten out 142 00:06:21,080 --> 00:06:23,040 Speaker 1: of ten, you can't miss it. I mean, I typed up, 143 00:06:23,040 --> 00:06:24,640 Speaker 1: and for good reason. I think we're gonna see a 144 00:06:24,640 --> 00:06:27,400 Speaker 1: great game. But the two components are what is the 145 00:06:27,720 --> 00:06:30,960 Speaker 1: average projected quality of these two teams? So what's the 146 00:06:30,960 --> 00:06:33,200 Speaker 1: average K forward rating of the two teams involved? And 147 00:06:33,240 --> 00:06:36,000 Speaker 1: then what's the average or excuse me, what's the projected 148 00:06:36,000 --> 00:06:39,080 Speaker 1: competitiveness of this game? So what's the projected k forward spread? 149 00:06:39,120 --> 00:06:41,080 Speaker 1: Of course Vegas has a spread, but I'm able to 150 00:06:41,080 --> 00:06:43,360 Speaker 1: generate my own spread for every game using my model, 151 00:06:43,480 --> 00:06:45,640 Speaker 1: and it's a weighted average of those two things. I 152 00:06:45,680 --> 00:06:48,680 Speaker 1: put more weight on the quality than the competitiveness, and 153 00:06:48,720 --> 00:06:51,560 Speaker 1: so that you know, tends to get power for teams 154 00:06:51,600 --> 00:06:53,880 Speaker 1: and matchups a little bit higher up, which draws the 155 00:06:53,920 --> 00:06:56,040 Speaker 1: ire of some G six fans out there, which hey, 156 00:06:56,040 --> 00:06:58,880 Speaker 1: I get, But I'm looking at those two things. What's 157 00:06:58,920 --> 00:07:01,719 Speaker 1: your average per projected quality of the teams? And what's 158 00:07:01,760 --> 00:07:05,120 Speaker 1: the projected competitiveness the spread of that game. Merge those 159 00:07:05,120 --> 00:07:07,200 Speaker 1: two things together in a weighted fashion. You get a 160 00:07:07,240 --> 00:07:09,840 Speaker 1: score of zero to ten for every single college football game, 161 00:07:09,880 --> 00:07:12,320 Speaker 1: and then I list those in the graphic on Fridays 162 00:07:12,480 --> 00:07:14,360 Speaker 1: based on the window. So I have a you know, 163 00:07:14,400 --> 00:07:16,640 Speaker 1: the top ten games of the early window, the top 164 00:07:16,640 --> 00:07:18,720 Speaker 1: ten games of the afternoon window, the top ten games 165 00:07:18,720 --> 00:07:21,120 Speaker 1: of primetime, and then the prop top ten games of 166 00:07:21,520 --> 00:07:24,200 Speaker 1: you know, late night and or weeknight games. And people 167 00:07:24,240 --> 00:07:25,200 Speaker 1: seem to really enjoy it. 168 00:07:26,000 --> 00:07:28,000 Speaker 2: All right, Let's go to another little project that you 169 00:07:28,040 --> 00:07:30,280 Speaker 2: put out there, which I absolutely love. You do it 170 00:07:30,280 --> 00:07:32,880 Speaker 2: throughout the season and updated weekly, which is the most 171 00:07:32,880 --> 00:07:36,920 Speaker 2: deserving college football Playoff bracket. And this is a concept 172 00:07:36,920 --> 00:07:39,400 Speaker 2: that college football fans, particularly those in the SEC, love 173 00:07:39,440 --> 00:07:41,520 Speaker 2: to argue about all the time. There's a huge difference 174 00:07:41,680 --> 00:07:43,680 Speaker 2: between how good you are and how good you could be. 175 00:07:43,760 --> 00:07:46,280 Speaker 2: So you go look at team composite rating from twenty 176 00:07:46,280 --> 00:07:49,280 Speaker 2: four to seven. That kind of evaluates the overall you know, 177 00:07:49,320 --> 00:07:52,400 Speaker 2: recruiting heft of your squad, how good you've been, you know, 178 00:07:52,480 --> 00:07:55,720 Speaker 2: considering the historical production of your coaching staff, of the 179 00:07:55,760 --> 00:07:59,800 Speaker 2: scheme you're running versus what you've actually accomplished purely what 180 00:07:59,840 --> 00:08:02,200 Speaker 2: you done on the field with the opportunities that have 181 00:08:02,240 --> 00:08:04,440 Speaker 2: been laid in front of you for that current season. 182 00:08:04,760 --> 00:08:07,440 Speaker 2: So when you release this bracket each week, what makes 183 00:08:07,480 --> 00:08:10,080 Speaker 2: these teams most deserving in your mind? And if you 184 00:08:10,080 --> 00:08:12,160 Speaker 2: could share a little bit, what's underneath the hood, what 185 00:08:12,320 --> 00:08:16,000 Speaker 2: powers this numbers Because there's always these arguments from fans, 186 00:08:16,000 --> 00:08:18,680 Speaker 2: whether it's casuals or ones who have skin in the 187 00:08:18,680 --> 00:08:21,440 Speaker 2: game with a particular team or conference that obviously want 188 00:08:21,440 --> 00:08:24,720 Speaker 2: to combat and fight and argue over a most deserving ranking, 189 00:08:24,760 --> 00:08:27,760 Speaker 2: but certainly something that even the College Football Committee in 190 00:08:27,800 --> 00:08:30,040 Speaker 2: real life they have to be at odds with because 191 00:08:30,080 --> 00:08:33,440 Speaker 2: there is that competitive balance between this team in a 192 00:08:33,520 --> 00:08:36,079 Speaker 2: vacuum is better than this other team, but they haven't 193 00:08:36,080 --> 00:08:37,360 Speaker 2: shown it on the field just yet. 194 00:08:38,000 --> 00:08:41,440 Speaker 1: Yeah, this is my favorite content that I produce, Mike, 195 00:08:41,600 --> 00:08:44,199 Speaker 1: is what I call the most Deserving Rankings, And yes, 196 00:08:44,360 --> 00:08:45,800 Speaker 1: I present it in a couple different ways. 197 00:08:45,800 --> 00:08:47,800 Speaker 4: One of them is the bracket that you referenced. 198 00:08:48,120 --> 00:08:51,080 Speaker 1: This content gets updated every single Tuesday on the website 199 00:08:51,120 --> 00:08:54,040 Speaker 1: and on X and it gets more informative the later 200 00:08:54,080 --> 00:08:56,319 Speaker 1: in the season we get, Like the most Deserving rankings, 201 00:08:56,920 --> 00:09:00,720 Speaker 1: they're very informative. Once teams up played six, seven, eight, nine, ten, 202 00:09:00,840 --> 00:09:03,080 Speaker 1: twelve games, like right now and early in the season, 203 00:09:03,160 --> 00:09:05,400 Speaker 1: you're gonna get some wonky results because it's all based 204 00:09:05,440 --> 00:09:08,480 Speaker 1: on what have you achieved to date on the field 205 00:09:08,480 --> 00:09:11,200 Speaker 1: this year. So, like after this weekend, either Texas or 206 00:09:11,240 --> 00:09:13,720 Speaker 1: Ohio State is gonna be number one of the most 207 00:09:13,760 --> 00:09:16,080 Speaker 1: Deserving rankings because they have beaten you know, either a 208 00:09:16,080 --> 00:09:18,839 Speaker 1: power rated number one or number two three team, right, 209 00:09:18,920 --> 00:09:20,840 Speaker 1: so they're gonna be number one. The loser of that 210 00:09:20,880 --> 00:09:22,800 Speaker 1: game is gonna be very, very far down in the rankings. 211 00:09:22,880 --> 00:09:25,079 Speaker 1: I mean, like in the sixties or the fifties, the sixties, 212 00:09:25,160 --> 00:09:27,600 Speaker 1: the seventies, because they're zero to one this season. Now, 213 00:09:27,640 --> 00:09:28,959 Speaker 1: of course they're not gonna be ranked. They're in the 214 00:09:29,000 --> 00:09:31,520 Speaker 1: ap pole. They're not this fifties, sixty, seventy the best 215 00:09:31,520 --> 00:09:33,520 Speaker 1: team in the country. But that's how the most Deserving 216 00:09:33,559 --> 00:09:34,120 Speaker 1: rankings work. 217 00:09:34,120 --> 00:09:37,040 Speaker 4: And you're right. It is funny on Twitter. 218 00:09:36,800 --> 00:09:39,360 Speaker 1: On x to see fan bases get so fired up, 219 00:09:39,480 --> 00:09:42,160 Speaker 1: but it is serious because the committee in real life, 220 00:09:42,160 --> 00:09:44,320 Speaker 1: how we determine the playoff, they need to be paying 221 00:09:44,320 --> 00:09:47,200 Speaker 1: attention to things, in my opinion, like the most Deserving rankings. 222 00:09:47,200 --> 00:09:50,120 Speaker 1: So I'll start with just the description of what it 223 00:09:50,240 --> 00:09:52,680 Speaker 1: is like, literally the definition that I put on the 224 00:09:52,679 --> 00:09:55,800 Speaker 1: graphic when I post it. So most deserving rankings are 225 00:09:56,080 --> 00:10:00,240 Speaker 1: results oriented or resume based measure of how now a 226 00:10:00,320 --> 00:10:04,760 Speaker 1: team has performed against its schedule, and I include scoring margin. 227 00:10:04,800 --> 00:10:07,360 Speaker 1: That's important to note. I'll come back to it relative 228 00:10:07,679 --> 00:10:11,080 Speaker 1: to how the number twelve power rated team would be 229 00:10:11,160 --> 00:10:14,360 Speaker 1: expected to perform against that same schedule. It is not 230 00:10:14,559 --> 00:10:17,320 Speaker 1: a power rating, So that's kind of a mouthful, and 231 00:10:17,360 --> 00:10:19,400 Speaker 1: there's a lot included in there. What does all of 232 00:10:19,400 --> 00:10:22,320 Speaker 1: that mean? What I want to know is how much 233 00:10:22,360 --> 00:10:25,760 Speaker 1: have you any team one through one, thirty six and fbs, 234 00:10:25,880 --> 00:10:29,760 Speaker 1: how much have you your team achieved against your schedule 235 00:10:30,360 --> 00:10:33,920 Speaker 1: win loss record wise relative to what would be expected 236 00:10:34,280 --> 00:10:36,960 Speaker 1: of the number twelve power rated team if they played 237 00:10:37,240 --> 00:10:40,079 Speaker 1: that schedule to date, so through week zero three, week one, 238 00:10:40,160 --> 00:10:41,439 Speaker 1: through the end of the regular season. 239 00:10:41,480 --> 00:10:42,000 Speaker 4: Whatever it is. 240 00:10:42,280 --> 00:10:44,520 Speaker 1: Why the number twelve power rated team? Because we have 241 00:10:44,520 --> 00:10:47,280 Speaker 1: a twelve team playoff, you have to have a denominator 242 00:10:47,480 --> 00:10:51,280 Speaker 1: to normalize every single team's achievement against their schedule relative 243 00:10:51,280 --> 00:10:53,320 Speaker 1: to expect it. You could pick the number twelve team, 244 00:10:53,320 --> 00:10:55,160 Speaker 1: You could pick the number twenty five team. You can 245 00:10:55,200 --> 00:10:57,880 Speaker 1: pick any number you want. I've picked number twelve because 246 00:10:57,920 --> 00:10:59,920 Speaker 1: we have a twelve team playoff. If the playoff expans, 247 00:11:00,080 --> 00:11:03,760 Speaker 1: I'll likely change my denominator reference point. So how much 248 00:11:03,800 --> 00:11:06,920 Speaker 1: have you achieved against your schedule relative to expected of 249 00:11:06,960 --> 00:11:09,760 Speaker 1: that number twelve team against that same schedule. That's essentially 250 00:11:09,760 --> 00:11:13,000 Speaker 1: what ESPN's strength of record is, which is included in 251 00:11:13,200 --> 00:11:15,920 Speaker 1: the team sheets that the committee looks at. Now, ESPN 252 00:11:16,000 --> 00:11:19,240 Speaker 1: strength of record uses the ESPNFPI as the engine. I'm 253 00:11:19,280 --> 00:11:21,360 Speaker 1: not going to get ato FPI. It's a polarizing topic. 254 00:11:21,679 --> 00:11:24,040 Speaker 1: I use the k Ford ratings as the engine of 255 00:11:24,160 --> 00:11:27,000 Speaker 1: power ratings as the engine for my most deserving. While 256 00:11:27,000 --> 00:11:29,960 Speaker 1: I think my most deserving rankings are slightly better than 257 00:11:30,000 --> 00:11:32,160 Speaker 1: the strength of record, which I think is an amazing concept. 258 00:11:32,160 --> 00:11:34,400 Speaker 1: I mean, that's how I built the most deserving. Why 259 00:11:34,440 --> 00:11:36,560 Speaker 1: I think mine's a little bit better is because I 260 00:11:36,640 --> 00:11:41,280 Speaker 1: incorporate scoring margin relative to expected. So, for example, let's 261 00:11:41,280 --> 00:11:43,000 Speaker 1: look at games that have already happened, and let's just 262 00:11:43,000 --> 00:11:46,240 Speaker 1: look at Iowa State in Kansas State. If Iowa State 263 00:11:46,720 --> 00:11:50,120 Speaker 1: beat Kansas State, as they did by three points, the 264 00:11:50,400 --> 00:11:52,360 Speaker 1: ESPN strength of record is going to be what it is. 265 00:11:52,480 --> 00:11:55,199 Speaker 1: Iowa State could have beat Kansas State by fifty points 266 00:11:55,360 --> 00:11:57,160 Speaker 1: and it would still be what it is ioa state 267 00:11:57,160 --> 00:11:59,360 Speaker 1: strength of record. Beating Kansas State by three and beating 268 00:11:59,400 --> 00:12:02,280 Speaker 1: Kansas State by is the exact same. In my Most 269 00:12:02,320 --> 00:12:05,439 Speaker 1: Deserving Iowa State, if they win by fifty they're gonna 270 00:12:05,440 --> 00:12:09,320 Speaker 1: have a better k Ford Most Deserving ranking or the 271 00:12:09,400 --> 00:12:12,000 Speaker 1: raw rating than if they beat Kansas State by three 272 00:12:12,040 --> 00:12:14,600 Speaker 1: points because you won that game by a larger margin. 273 00:12:14,679 --> 00:12:16,319 Speaker 1: The number twelve power ated team is going to be 274 00:12:16,360 --> 00:12:18,480 Speaker 1: expected to beat Kansas State by X number of points. 275 00:12:18,640 --> 00:12:20,840 Speaker 1: Iowa State exceeded that or fell short of it by 276 00:12:20,880 --> 00:12:24,520 Speaker 1: a certain number. By incorporating relative scoring margin, I can 277 00:12:24,559 --> 00:12:27,080 Speaker 1: adjust for that. So that's why I really like my 278 00:12:27,120 --> 00:12:29,920 Speaker 1: Most Deserving. That's exactly what it is. That's how it works. 279 00:12:29,960 --> 00:12:33,720 Speaker 1: And then to your point, based on the results that 280 00:12:33,760 --> 00:12:35,640 Speaker 1: I have to date, which right now is only nine 281 00:12:35,640 --> 00:12:38,160 Speaker 1: teams because only nine FBS teams played in Week zero, 282 00:12:38,400 --> 00:12:40,960 Speaker 1: we only have nine teams in the bracket. After Week one, 283 00:12:41,080 --> 00:12:43,280 Speaker 1: everybody's gonna have played at least one game, some will 284 00:12:43,280 --> 00:12:45,760 Speaker 1: have played two games. We're gonna have looking a much 285 00:12:45,760 --> 00:12:48,439 Speaker 1: different looking bracket. And also, as it relates to the 286 00:12:48,480 --> 00:12:51,760 Speaker 1: most deserving rankings, you will always see the number twelve 287 00:12:51,760 --> 00:12:55,080 Speaker 1: team is gonna have a raw most deserving rating of one, 288 00:12:55,320 --> 00:12:57,960 Speaker 1: because remember how much have they achieved against their schedule 289 00:12:58,160 --> 00:13:00,880 Speaker 1: relative to the number twelve poweraded team. The number twelve 290 00:13:00,920 --> 00:13:02,560 Speaker 1: team is always going to have a power or a 291 00:13:02,600 --> 00:13:05,280 Speaker 1: most deserving rating of one. The bottom FBS team one 292 00:13:05,360 --> 00:13:07,320 Speaker 1: thirty six is always going to be zero. The top 293 00:13:07,360 --> 00:13:10,680 Speaker 1: fluctuates because that is basically how much farther ahead, how 294 00:13:10,760 --> 00:13:13,200 Speaker 1: much more have you achieved than the number twelve team 295 00:13:13,200 --> 00:13:16,000 Speaker 1: against that schedule. But the number twelve is always going 296 00:13:16,080 --> 00:13:17,520 Speaker 1: to be one and the number one thirty six is 297 00:13:17,520 --> 00:13:19,080 Speaker 1: always going to be zero. That kind of gives you 298 00:13:19,120 --> 00:13:21,400 Speaker 1: the scale for how those numbers get presented. 299 00:13:22,280 --> 00:13:24,480 Speaker 2: Football season is here, folks, and we've got some huge 300 00:13:24,480 --> 00:13:27,640 Speaker 2: improvements and updates coming to the award winning Action Network app, 301 00:13:27,720 --> 00:13:30,160 Speaker 2: which we're very excited to discuss in the weeks ahead. 302 00:13:30,360 --> 00:13:32,440 Speaker 2: But here's the thing. You're going to miss out without 303 00:13:32,440 --> 00:13:34,680 Speaker 2: the free Action Network app because you'll you get to 304 00:13:34,679 --> 00:13:36,360 Speaker 2: see our picks as soon as we book them. That 305 00:13:36,400 --> 00:13:39,960 Speaker 2: means my picks Ducks, Stuckeyes, Colin Wilson, all the experts 306 00:13:40,080 --> 00:13:42,320 Speaker 2: here at Action Network you will miss out if you 307 00:13:42,360 --> 00:13:45,000 Speaker 2: have not downloaded and signed up for the Action Network app. 308 00:13:45,040 --> 00:13:47,240 Speaker 2: And here's the thing. It is free to sign up 309 00:13:47,240 --> 00:13:49,240 Speaker 2: by using the link in the episode description. So go 310 00:13:49,280 --> 00:13:51,320 Speaker 2: ahead and do that and then follow me. You can 311 00:13:51,320 --> 00:13:55,080 Speaker 2: search for Breeze, that's br Ees in the Action app. 312 00:13:55,240 --> 00:13:57,199 Speaker 2: My name will pop up, Mike Calibers. You go ahead 313 00:13:57,200 --> 00:14:00,400 Speaker 2: and follow, And every single second that I put in 314 00:14:00,440 --> 00:14:02,480 Speaker 2: a pick, it's going to go through the app, it's 315 00:14:02,480 --> 00:14:04,240 Speaker 2: going to pop right there on your phone. Give you 316 00:14:04,280 --> 00:14:07,040 Speaker 2: an alert. That way, you're always getting the best numbers possible. 317 00:14:07,200 --> 00:14:09,080 Speaker 2: And when I'm going to the window, cash and winners. 318 00:14:09,240 --> 00:14:11,800 Speaker 2: So were you all right? Kelly? Here we go. How 319 00:14:11,840 --> 00:14:15,680 Speaker 2: difficult is it to project teams in the preseason? Because 320 00:14:15,800 --> 00:14:19,800 Speaker 2: I absolutely love how quickly your model updates in real time. 321 00:14:19,880 --> 00:14:21,960 Speaker 2: You have all these different inputs throughout the course of 322 00:14:22,000 --> 00:14:25,360 Speaker 2: the season. Teams are rising, they're falling. But in the 323 00:14:25,400 --> 00:14:29,440 Speaker 2: transfer portal era, from a preseason perspective, how do you 324 00:14:29,560 --> 00:14:32,800 Speaker 2: calibrate these things? Because I, honestly, I have to imagine 325 00:14:32,800 --> 00:14:36,040 Speaker 2: that the preseason ratings can be so far off from 326 00:14:36,360 --> 00:14:38,600 Speaker 2: let's even say five or six years ago, when there 327 00:14:38,640 --> 00:14:41,080 Speaker 2: was more roster continuity, when you had a little bit 328 00:14:41,080 --> 00:14:43,200 Speaker 2: of a better idea the brand of football that you 329 00:14:43,200 --> 00:14:44,720 Speaker 2: were going to see from a team year to year. 330 00:14:45,080 --> 00:14:47,880 Speaker 2: And because it's moving so quickly, did you have to 331 00:14:47,960 --> 00:14:51,840 Speaker 2: make any alterations to how quickly new data was impacting 332 00:14:51,880 --> 00:14:55,000 Speaker 2: Where these teams get pulled Colin and Stucky they do 333 00:14:55,040 --> 00:14:57,400 Speaker 2: their own power rankings. They talk about this a lot. Oh, 334 00:14:57,440 --> 00:14:59,080 Speaker 2: all of a sudden, I have to upgrade this team 335 00:14:59,120 --> 00:15:01,680 Speaker 2: by ten or fifteen slots, where maybe a few years 336 00:15:01,680 --> 00:15:04,200 Speaker 2: ago it would have been a slower burn to get 337 00:15:04,200 --> 00:15:06,960 Speaker 2: them on the move, But now I have to undervalue 338 00:15:07,000 --> 00:15:10,000 Speaker 2: a bit of the preseason value in the rankings and 339 00:15:10,040 --> 00:15:12,120 Speaker 2: put more eggs in the basket of what I'm seeing 340 00:15:12,160 --> 00:15:14,400 Speaker 2: on the field in twenty twenty five, Walk me through 341 00:15:14,440 --> 00:15:17,160 Speaker 2: that kind of push and pull between those two data points. 342 00:15:17,560 --> 00:15:19,160 Speaker 4: Yeah, that's exactly right, Mike. 343 00:15:19,280 --> 00:15:21,320 Speaker 1: And so to your point, you know, when I started 344 00:15:21,360 --> 00:15:24,880 Speaker 1: doing this publicly back in twenty nineteen, the way that 345 00:15:24,920 --> 00:15:29,480 Speaker 1: you calculated preseason power ratings was pretty consistent year over year. 346 00:15:29,520 --> 00:15:31,600 Speaker 1: And you're looking at your returning production. You were looking 347 00:15:31,640 --> 00:15:35,360 Speaker 1: at how good has this program been in the last two, three, four, 348 00:15:35,600 --> 00:15:39,040 Speaker 1: five years, how well have they recruited in the last two, three, 349 00:15:39,120 --> 00:15:42,200 Speaker 1: four years from the high school level. Now with the 350 00:15:42,200 --> 00:15:44,560 Speaker 1: way rosters are constructed and the way they turn over, 351 00:15:44,800 --> 00:15:48,200 Speaker 1: and we've seen this really with the increased usage of 352 00:15:48,200 --> 00:15:50,280 Speaker 1: the transferportal, not the advent of the transfer portal, but 353 00:15:50,320 --> 00:15:53,640 Speaker 1: the increased usage of it. Returning production still matters, and 354 00:15:53,680 --> 00:15:56,720 Speaker 1: it matters greatly, and I incorporate transfers into the returning 355 00:15:56,760 --> 00:15:59,320 Speaker 1: production portion. And of course you have to have scalers 356 00:15:59,320 --> 00:16:02,680 Speaker 1: and multiplayers, right because if a player's come into Texas 357 00:16:02,760 --> 00:16:06,840 Speaker 1: A and M from Alabama, Okay, that's pretty we can 358 00:16:06,920 --> 00:16:09,120 Speaker 1: expect that the production that had Albama's going to translate 359 00:16:09,120 --> 00:16:12,040 Speaker 1: pretty well. If they're coming from Miami of Ohio and 360 00:16:12,080 --> 00:16:13,920 Speaker 1: they had a great season, you're probably gonna downgrade what 361 00:16:13,960 --> 00:16:16,240 Speaker 1: you're epecting because you're stepping up in competition. Similarly, if 362 00:16:16,240 --> 00:16:17,960 Speaker 1: you go the other way, you're gonna upgrade if somebody's 363 00:16:17,960 --> 00:16:21,200 Speaker 1: going from A and M to Miami, Ohio. So that's 364 00:16:21,240 --> 00:16:23,840 Speaker 1: all been a part of this. The importance of coaches 365 00:16:24,400 --> 00:16:27,360 Speaker 1: now in coaching continuity. That's something I've incorporated into the 366 00:16:27,360 --> 00:16:29,640 Speaker 1: model for the first time here and I think is 367 00:16:29,680 --> 00:16:31,760 Speaker 1: proving to be and we'll see as we continue to 368 00:16:31,960 --> 00:16:34,000 Speaker 1: get years of data on this as I move forward, 369 00:16:34,040 --> 00:16:37,520 Speaker 1: but I think that's becoming increasingly important. What's becoming decreasingly 370 00:16:37,560 --> 00:16:40,520 Speaker 1: important is how good was your team two, three, four 371 00:16:40,600 --> 00:16:43,960 Speaker 1: years ago? Decreasingly important? How well are you recruiting at 372 00:16:44,000 --> 00:16:46,680 Speaker 1: the high school level. It still matters, but it matters 373 00:16:46,760 --> 00:16:49,200 Speaker 1: less now because a lot of the work and bringing 374 00:16:49,240 --> 00:16:51,920 Speaker 1: on new players in your team is through the transfer portal, 375 00:16:52,160 --> 00:16:54,080 Speaker 1: and you can flip a team so easily now. I mean, 376 00:16:54,080 --> 00:16:56,680 Speaker 1: you look at Purdue this year. They've got like seventy 377 00:16:56,760 --> 00:16:59,360 Speaker 1: new guys. I don't really care how good Purdue was 378 00:16:59,440 --> 00:17:02,080 Speaker 1: last year or or two years ago, because literally no 379 00:17:02,160 --> 00:17:04,359 Speaker 1: one was it is still there that was on those teams. 380 00:17:04,359 --> 00:17:07,639 Speaker 1: So the preseason look is much much different. And so 381 00:17:07,720 --> 00:17:10,280 Speaker 1: the final point that you add a question you asked there, 382 00:17:10,720 --> 00:17:14,160 Speaker 1: how quickly do we phase out preseason power ratings. It's 383 00:17:14,200 --> 00:17:16,399 Speaker 1: something you gotta make a decision, and that's what you 384 00:17:16,440 --> 00:17:19,919 Speaker 1: stick with. But this is something that I've accelerated in 385 00:17:20,040 --> 00:17:23,239 Speaker 1: the most recent years. Here I will see how it 386 00:17:23,280 --> 00:17:25,560 Speaker 1: goes this year. At the end of the season, I'll 387 00:17:25,640 --> 00:17:28,600 Speaker 1: kind of back test how would the model performed how 388 00:17:28,600 --> 00:17:30,520 Speaker 1: it did based on the decisions I made. How would 389 00:17:30,520 --> 00:17:34,320 Speaker 1: it have performed if I phased out preseason waits more quickly. 390 00:17:34,440 --> 00:17:36,560 Speaker 1: How would it have performed if I phase them out 391 00:17:36,560 --> 00:17:38,639 Speaker 1: more slowly just to kind of get a good idea. 392 00:17:38,640 --> 00:17:40,680 Speaker 1: And I do this every single summer when you're back 393 00:17:40,680 --> 00:17:42,359 Speaker 1: testing and looking at what tweets can we make for 394 00:17:42,400 --> 00:17:45,280 Speaker 1: the upcoming year. So, yes, how good your program's been 395 00:17:45,280 --> 00:17:49,440 Speaker 1: in recent years matters less. Coaches matter more, and preseason 396 00:17:49,440 --> 00:17:52,960 Speaker 1: components get phased out more quickly today than they did. 397 00:17:53,040 --> 00:17:55,359 Speaker 1: You know in twenty nineteen or twenty twenty one. I 398 00:17:55,320 --> 00:17:58,439 Speaker 1: skipped the twenty twenty season. That was just an outlier 399 00:17:58,520 --> 00:18:00,920 Speaker 1: year with COVID effects. But yes, it's definitely change, Mike. 400 00:18:01,560 --> 00:18:04,639 Speaker 2: All right, let's run through the week one docket, lining 401 00:18:04,720 --> 00:18:07,840 Speaker 2: up the Vegas line versus your ratings. Which games do 402 00:18:07,840 --> 00:18:10,639 Speaker 2: you think better should consider playing because your rankings have 403 00:18:10,720 --> 00:18:11,960 Speaker 2: spotted a discrepancy. 404 00:18:12,359 --> 00:18:14,800 Speaker 1: Yeah, as I'll give two that I think there might 405 00:18:14,840 --> 00:18:17,960 Speaker 1: be some value in. And it's interesting because they're both 406 00:18:17,960 --> 00:18:20,119 Speaker 1: they both involved teams that played last week, and they 407 00:18:20,160 --> 00:18:22,639 Speaker 1: played each other. So I see the Vegas line for 408 00:18:22,760 --> 00:18:25,920 Speaker 1: Kansas State at home against North Dakota is twenty five. 409 00:18:26,680 --> 00:18:30,199 Speaker 1: I've got Kansas State by a few more points than that. 410 00:18:30,840 --> 00:18:33,239 Speaker 1: If I look at my expected my win expectancy for 411 00:18:33,280 --> 00:18:36,240 Speaker 1: them in this game, I've got it at ninety nine percent, 412 00:18:36,320 --> 00:18:38,200 Speaker 1: which that comes out to more than a twenty five 413 00:18:38,240 --> 00:18:41,400 Speaker 1: point spread. So I think Kansas State at home bouncing 414 00:18:41,440 --> 00:18:44,159 Speaker 1: back from Iowa State. Now somebody to consider. They're coming 415 00:18:44,200 --> 00:18:47,640 Speaker 1: back from Ireland, right, my model is not explicitly accounting 416 00:18:47,680 --> 00:18:50,000 Speaker 1: for that travel that they have. North Dakota could have 417 00:18:50,040 --> 00:18:52,440 Speaker 1: beaten them to Manhattan if they really wanted to, coming 418 00:18:52,560 --> 00:18:54,399 Speaker 1: from just a couple of states away, and they're coming 419 00:18:54,720 --> 00:18:58,000 Speaker 1: the Wildcats coming from across the pond. So that's somebody 420 00:18:58,000 --> 00:19:00,200 Speaker 1: to consider. But my model still really likes Kansas State 421 00:19:00,240 --> 00:19:01,879 Speaker 1: to bounce back. And then the other game that the 422 00:19:01,920 --> 00:19:05,560 Speaker 1: model really sees a discrepancy, and again same travel considerations, 423 00:19:05,840 --> 00:19:08,280 Speaker 1: Iowa State minus fourteen and a half. I've got a 424 00:19:08,359 --> 00:19:10,720 Speaker 1: ninety four percent win expectancy here, So my model does 425 00:19:10,760 --> 00:19:13,160 Speaker 1: think Iowa State wins this game by you know, seventeen 426 00:19:13,200 --> 00:19:14,959 Speaker 1: points plus. So I think there is a little bit 427 00:19:15,000 --> 00:19:17,480 Speaker 1: of value there. Some other ones that the model called 428 00:19:17,520 --> 00:19:20,720 Speaker 1: out that I would say consider staying away for a 429 00:19:20,720 --> 00:19:23,360 Speaker 1: couple of reasons. Florida minus forty five and a half 430 00:19:23,440 --> 00:19:27,760 Speaker 1: versus LIU, that's a really really big number. They can 431 00:19:27,840 --> 00:19:29,600 Speaker 1: cover it, but you never really know how it goes 432 00:19:29,640 --> 00:19:31,480 Speaker 1: when the backups get in there, and Week one is 433 00:19:31,480 --> 00:19:34,200 Speaker 1: always tough with especially when you're projecting FCS teams. That's 434 00:19:34,240 --> 00:19:37,440 Speaker 1: not my primary area. Ball State plus eighteen and a 435 00:19:37,480 --> 00:19:38,480 Speaker 1: half at Purdue. 436 00:19:38,720 --> 00:19:40,040 Speaker 4: This is a completely new Perdue team. 437 00:19:40,080 --> 00:19:43,080 Speaker 1: As I said, so Perdue could win by more than 438 00:19:43,119 --> 00:19:44,720 Speaker 1: eighteen and a half and then I'd feel bad about 439 00:19:44,720 --> 00:19:46,639 Speaker 1: saying this. But as the model looks at what we 440 00:19:46,680 --> 00:19:49,280 Speaker 1: know about this bary otom team, ball State eighteen and 441 00:19:49,280 --> 00:19:52,000 Speaker 1: a half might be some value. Iowa minus thirty eight 442 00:19:52,040 --> 00:19:54,719 Speaker 1: at home against you Albany. Thirty eight is a big number, period, 443 00:19:54,880 --> 00:19:57,200 Speaker 1: especially for Iowa team. I do expect their offense to 444 00:19:57,240 --> 00:19:59,640 Speaker 1: be better this year, but let's see how many times 445 00:19:59,680 --> 00:20:01,800 Speaker 1: Iowa was able to score thirty eight points this season, 446 00:20:01,800 --> 00:20:03,520 Speaker 1: and that's what they have to do here at a minimum. 447 00:20:03,560 --> 00:20:05,960 Speaker 1: If you're a true sicko, think about Georgia State plus 448 00:20:06,000 --> 00:20:08,320 Speaker 1: thirty five and a half at Old Miss Stonybrook plus 449 00:20:08,320 --> 00:20:11,200 Speaker 1: eighteen and a half at San Diego State, SMU minus 450 00:20:11,240 --> 00:20:13,040 Speaker 1: forty eight and a half versus East Texas A and 451 00:20:13,200 --> 00:20:15,720 Speaker 1: M and Hawaii plus seventeen and a half at Arizona. 452 00:20:15,760 --> 00:20:17,719 Speaker 1: Those are some other ones that the model kind of 453 00:20:17,880 --> 00:20:21,680 Speaker 1: called out, but again, those are really sicko plays if 454 00:20:21,720 --> 00:20:22,800 Speaker 1: you're trying to dive into. 455 00:20:22,600 --> 00:20:24,760 Speaker 2: It, all right, If I could pause and go back 456 00:20:24,800 --> 00:20:27,240 Speaker 2: to two games in particular, you talked about k State 457 00:20:27,320 --> 00:20:30,400 Speaker 2: at the top of this segment. How do your ratings 458 00:20:30,520 --> 00:20:33,439 Speaker 2: fold in injury concerns or do they just kind of 459 00:20:33,440 --> 00:20:35,640 Speaker 2: gloss over them and say, you know, in the wash 460 00:20:35,680 --> 00:20:38,560 Speaker 2: over the course of the season, over hundreds and thousands 461 00:20:38,560 --> 00:20:41,040 Speaker 2: of games, literally, you know, it's probably for the best 462 00:20:41,119 --> 00:20:43,600 Speaker 2: not to have that noise inside of your model. Because 463 00:20:43,640 --> 00:20:46,040 Speaker 2: Dylan Edwards he may go, he may not. He injured 464 00:20:46,080 --> 00:20:47,760 Speaker 2: his ankle on the very first play when he was 465 00:20:47,800 --> 00:20:51,280 Speaker 2: receiving a punt in that Pharmo Gedden game. If he plays, 466 00:20:51,320 --> 00:20:53,639 Speaker 2: obviously he could be an offensive engine for the Wildcats. 467 00:20:53,640 --> 00:20:55,440 Speaker 2: If he doesn't, are they going to be stuck in 468 00:20:55,480 --> 00:20:56,879 Speaker 2: the mud for a bit of the game as they 469 00:20:56,920 --> 00:20:59,159 Speaker 2: were against Iowa State. What are your thoughts on that 470 00:20:59,240 --> 00:21:01,200 Speaker 2: element before I get my second question. 471 00:21:01,320 --> 00:21:04,439 Speaker 1: Yeah, really good question and something that every better needs 472 00:21:04,480 --> 00:21:06,480 Speaker 1: to keep in mind, right, Like, my numbers are a 473 00:21:06,520 --> 00:21:08,679 Speaker 1: starting point, and this is the type of analysis that 474 00:21:08,720 --> 00:21:11,400 Speaker 1: you have to do for every single game that's out 475 00:21:11,400 --> 00:21:15,480 Speaker 1: there on the card. The answer, Mike to your question is, 476 00:21:15,480 --> 00:21:17,560 Speaker 1: if it's a quarterback, I go in and make a 477 00:21:17,560 --> 00:21:22,280 Speaker 1: manual adjustment. If it's not a quarterback, I typically leave 478 00:21:22,320 --> 00:21:24,639 Speaker 1: it alone. And what happens is if the player is 479 00:21:24,680 --> 00:21:27,159 Speaker 1: out for an extended period of time, the model is 480 00:21:27,240 --> 00:21:31,480 Speaker 1: implicitly capturing that missing production. It's like, oh wow, they're 481 00:21:31,520 --> 00:21:35,040 Speaker 1: not getting much production here at this position from Kansas State. 482 00:21:35,520 --> 00:21:36,479 Speaker 4: There's something going on. 483 00:21:36,600 --> 00:21:39,520 Speaker 1: So the model's gonna learn over time that a player 484 00:21:39,600 --> 00:21:41,920 Speaker 1: is missing if it's a quarterback, because of the role 485 00:21:42,000 --> 00:21:44,399 Speaker 1: they play in the sport of football, not just college, 486 00:21:44,400 --> 00:21:46,399 Speaker 1: but in the sport of football. I go in and 487 00:21:46,440 --> 00:21:48,840 Speaker 1: make a manual adjustment to make sure that the model 488 00:21:48,920 --> 00:21:51,399 Speaker 1: knows which quarterback is playing, and we can expect that 489 00:21:51,520 --> 00:21:54,200 Speaker 1: level of play and competency from a team in any 490 00:21:54,200 --> 00:21:56,639 Speaker 1: given week. But outside of quarterback, I don't make the 491 00:21:56,640 --> 00:21:59,000 Speaker 1: manual adjustments, and we count on the model to implicitly 492 00:21:59,040 --> 00:22:01,119 Speaker 1: capture it if that player ends up being out for 493 00:22:01,160 --> 00:22:02,200 Speaker 1: an extended period of time. 494 00:22:02,200 --> 00:22:04,560 Speaker 4: But great, great question, Really important for people to know. 495 00:22:05,480 --> 00:22:09,080 Speaker 2: Second part, in terms of manual adjustments, you mentioned earlier 496 00:22:09,160 --> 00:22:12,000 Speaker 2: that when you look at a player moving from a Miami, 497 00:22:12,080 --> 00:22:14,560 Speaker 2: Ohio to Texas A and M that you're gonna have 498 00:22:14,560 --> 00:22:16,920 Speaker 2: to take that into account. Let's move even further down 499 00:22:17,040 --> 00:22:20,040 Speaker 2: the food chain. Mark Gronowski comes in from South Dakota 500 00:22:20,080 --> 00:22:23,480 Speaker 2: State but potentially could be a massive difference maker for 501 00:22:23,520 --> 00:22:25,919 Speaker 2: the Iowa passing attack. Did you have to go in 502 00:22:25,960 --> 00:22:28,399 Speaker 2: and make a manual adjustment visa B how you usually 503 00:22:28,480 --> 00:22:30,919 Speaker 2: view FCS transfers or you're just going to let it 504 00:22:30,960 --> 00:22:33,760 Speaker 2: play out and then see if the model can recalibrate itself. 505 00:22:33,800 --> 00:22:35,560 Speaker 2: Here and let's call it the first three weeks of 506 00:22:35,560 --> 00:22:36,679 Speaker 2: the twenty twenty five season. 507 00:22:37,000 --> 00:22:39,840 Speaker 1: So what's interesting about this is, yes, we do have 508 00:22:39,920 --> 00:22:43,600 Speaker 1: kind of like baseline scalers and modifiers if you will, 509 00:22:43,840 --> 00:22:47,239 Speaker 1: from FCS to G six to power four, power two, 510 00:22:47,320 --> 00:22:50,359 Speaker 1: whatever you want to call it. Over time, we have 511 00:22:50,400 --> 00:22:54,359 Speaker 1: his historical data set for how do these pieces tend 512 00:22:54,359 --> 00:22:57,080 Speaker 1: to fit and perform at their new landing spot. It's 513 00:22:57,119 --> 00:22:59,720 Speaker 1: always interesting though, when you have players coming from you 514 00:23:00,119 --> 00:23:03,920 Speaker 1: or highly successful FCS or G six programs like that 515 00:23:04,480 --> 00:23:08,800 Speaker 1: doesn't always explicitly get captured. So no, I didn't make 516 00:23:08,800 --> 00:23:13,080 Speaker 1: any like special exceptions or manual overrides in this case. 517 00:23:13,359 --> 00:23:17,080 Speaker 1: But the model, I've been doing this for long enough 518 00:23:17,119 --> 00:23:19,439 Speaker 1: now that the data that I have built up on 519 00:23:19,480 --> 00:23:22,800 Speaker 1: the back end is pretty it's pretty intelligent, and it's 520 00:23:22,840 --> 00:23:25,280 Speaker 1: able to really sift through a lot of the noise. 521 00:23:25,320 --> 00:23:27,080 Speaker 1: That I use that term earlier, and it's a great 522 00:23:27,160 --> 00:23:29,080 Speaker 1: term that we use in modeling, like we're able to 523 00:23:29,080 --> 00:23:31,280 Speaker 1: sift through some of that noise and really hone in 524 00:23:31,320 --> 00:23:34,920 Speaker 1: on what's important. So I'm very comfortable with the scalers 525 00:23:34,920 --> 00:23:37,560 Speaker 1: and the modifiers that I have in place for returning 526 00:23:37,600 --> 00:23:41,480 Speaker 1: production as players are transferring from one school to another, 527 00:23:42,080 --> 00:23:45,320 Speaker 1: transcending levels of competition, if you will. 528 00:23:45,400 --> 00:23:48,000 Speaker 2: Let's close out with two high level questions. The first 529 00:23:48,280 --> 00:23:51,080 Speaker 2: in terms of where your model is rating them VISA VI. 530 00:23:51,600 --> 00:23:55,080 Speaker 2: The talking heads, the public, the casuals. This is really 531 00:23:55,160 --> 00:23:57,480 Speaker 2: the end of talking season. This is the last opportunity 532 00:23:57,640 --> 00:23:59,760 Speaker 2: for everyone to throw their hot takes out. Who's winning 533 00:23:59,760 --> 00:24:01,960 Speaker 2: the conference, who's winging the Heisman Trophy, Who's going to 534 00:24:02,000 --> 00:24:05,160 Speaker 2: the College Football playoff? Two teams? Give me one for each. 535 00:24:05,359 --> 00:24:09,639 Speaker 2: Most overrated in your opinion VISAVI your ratings, and most underrated. 536 00:24:10,359 --> 00:24:13,440 Speaker 1: Yeah, let me really quickly hear pull up the AP Pole, 537 00:24:13,440 --> 00:24:15,040 Speaker 1: because that's how I'm going to kind of determine this. 538 00:24:15,119 --> 00:24:16,639 Speaker 1: I think the one that stands out to me and 539 00:24:16,680 --> 00:24:19,200 Speaker 1: that fans have already gotten mad at me about Illinois. 540 00:24:19,359 --> 00:24:21,199 Speaker 4: I know Brett Bielomo is going to be all over me. 541 00:24:21,320 --> 00:24:21,919 Speaker 4: He likes to do that. 542 00:24:22,000 --> 00:24:24,280 Speaker 1: I guess with people. I'm probably not big enough for 543 00:24:24,320 --> 00:24:27,000 Speaker 1: Brett to know about me. But Illinois number twelve in 544 00:24:27,040 --> 00:24:30,280 Speaker 1: the preseason AP Pole if I'm not mistaken here, and 545 00:24:30,320 --> 00:24:32,879 Speaker 1: I've got them number thirty in my power rating, So 546 00:24:32,920 --> 00:24:36,959 Speaker 1: that's that's a pretty significant difference, I'll say. And then 547 00:24:37,000 --> 00:24:39,680 Speaker 1: if we go the other way, man at the top, 548 00:24:39,760 --> 00:24:43,360 Speaker 1: we're looking pretty good. Okay, wait, here's here's one Texas 549 00:24:43,400 --> 00:24:46,720 Speaker 1: A and them. I have number ten. The AP pole 550 00:24:46,760 --> 00:24:50,760 Speaker 1: has them at number nineteen. Now it's always interesting because 551 00:24:51,040 --> 00:24:53,920 Speaker 1: especially schools in the SEC, when I say a team 552 00:24:53,960 --> 00:24:55,880 Speaker 1: is power at number ten or like, for me, Florida's 553 00:24:55,960 --> 00:24:59,840 Speaker 1: number eleven. But the projected record of these teams is 554 00:25:00,440 --> 00:25:03,480 Speaker 1: eight and four, seven and five, So I don't expect 555 00:25:03,560 --> 00:25:06,280 Speaker 1: them to be ranked in the AP pole number ten 556 00:25:06,359 --> 00:25:08,199 Speaker 1: or number eleven at the end of the year. My 557 00:25:08,280 --> 00:25:10,879 Speaker 1: most deserving is gonna account for the level of difficulty 558 00:25:10,920 --> 00:25:13,520 Speaker 1: that they played in their schedule, but their record is 559 00:25:13,560 --> 00:25:15,159 Speaker 1: not going to indicate that they're a top ten or 560 00:25:15,160 --> 00:25:17,359 Speaker 1: top fifteen team at the end of the year. But 561 00:25:17,520 --> 00:25:20,560 Speaker 1: right now, their talent can posit and what we're looking 562 00:25:20,560 --> 00:25:23,280 Speaker 1: at in terms of predictive modeling says that A and 563 00:25:23,400 --> 00:25:24,800 Speaker 1: M in Florida to a electric sent I think the 564 00:25:24,800 --> 00:25:26,920 Speaker 1: APS and fifteen, but Texas A and M at nineteen 565 00:25:26,920 --> 00:25:28,440 Speaker 1: and the AP I've got a number ten. That's a 566 00:25:28,480 --> 00:25:31,600 Speaker 1: pretty big discrepancy at the top of the either power 567 00:25:31,640 --> 00:25:33,840 Speaker 1: rating or AP pole, which at this time basically is 568 00:25:33,840 --> 00:25:35,280 Speaker 1: a power rating. Because you haven't played any games, how 569 00:25:35,280 --> 00:25:37,320 Speaker 1: are you ranking anybody? So that's what I'll say, A 570 00:25:37,400 --> 00:25:40,480 Speaker 1: and M. I think the public is underrating and perhaps 571 00:25:40,560 --> 00:25:44,280 Speaker 1: Illinois from a talent standpoint, might be a bit overrated 572 00:25:44,520 --> 00:25:45,479 Speaker 1: in the eyes of the public. 573 00:25:46,359 --> 00:25:48,639 Speaker 2: Now you've been on this data journey both publicly and 574 00:25:48,680 --> 00:25:51,680 Speaker 2: now you know it excuse me, previously privately, but now 575 00:25:51,720 --> 00:25:53,480 Speaker 2: it's public. You can go to your website, you can 576 00:25:53,520 --> 00:25:56,160 Speaker 2: see everything that you're putting on social media. How much 577 00:25:56,200 --> 00:25:58,560 Speaker 2: have you been surprised by some of the findings, you know, 578 00:25:58,640 --> 00:26:02,199 Speaker 2: really digging down doing the work in you know, in 579 00:26:02,240 --> 00:26:04,760 Speaker 2: the solo bunker almost then you have to present it 580 00:26:04,760 --> 00:26:08,800 Speaker 2: to the world, whether it's data modeling, predictive analytics, essentially 581 00:26:08,800 --> 00:26:11,719 Speaker 2: creating and fine tuning these k forward ratings. What has 582 00:26:11,760 --> 00:26:14,120 Speaker 2: surprised you the most from this data journey? 583 00:26:14,440 --> 00:26:17,200 Speaker 1: Yeah, I think I'll give a tongue in cheek answer, 584 00:26:17,240 --> 00:26:21,240 Speaker 1: and then my real one people on social media like, 585 00:26:22,240 --> 00:26:24,919 Speaker 1: it's crazy if your team is not number one in 586 00:26:24,960 --> 00:26:27,280 Speaker 1: a list that I put out, they're not happy. So like, 587 00:26:27,480 --> 00:26:29,880 Speaker 1: I make one fan base happy every time I put 588 00:26:29,880 --> 00:26:32,440 Speaker 1: out a list, and every other fan base is mad 589 00:26:32,680 --> 00:26:35,000 Speaker 1: and they think I'm biased. They think I hate their team. 590 00:26:35,200 --> 00:26:37,200 Speaker 1: We've had a podcast a few years ago called we 591 00:26:37,240 --> 00:26:39,600 Speaker 1: Hate Your Team because that's like the that's the most 592 00:26:39,600 --> 00:26:41,960 Speaker 1: common feedback I got, especially early on Kelly, why do 593 00:26:42,000 --> 00:26:44,960 Speaker 1: you hate my team? I've taken an objective approach with 594 00:26:45,000 --> 00:26:47,200 Speaker 1: the model to all one hundred and thirty six now 595 00:26:47,320 --> 00:26:51,520 Speaker 1: FBS teams, and you know, I try to objectively quantify 596 00:26:51,880 --> 00:26:54,000 Speaker 1: what the data says about these teams, and then I 597 00:26:54,040 --> 00:26:56,040 Speaker 1: just present the results and that's what I post, and 598 00:26:56,480 --> 00:26:59,200 Speaker 1: very very rarely do I get into like my opinion 599 00:26:59,320 --> 00:27:00,920 Speaker 1: of it. I just kind to post what the model 600 00:27:00,960 --> 00:27:04,080 Speaker 1: spitting out, but from a serious note. I think the 601 00:27:04,880 --> 00:27:07,159 Speaker 1: biggest thing, and maybe it's not a surprise, but the 602 00:27:07,200 --> 00:27:08,760 Speaker 1: biggest takeaway I have as. 603 00:27:08,640 --> 00:27:09,240 Speaker 4: Of right now. 604 00:27:09,280 --> 00:27:12,920 Speaker 1: We talked earlier about the preseason component I've really had 605 00:27:12,920 --> 00:27:16,720 Speaker 1: to modify. How do you go about trying to accurately 606 00:27:17,040 --> 00:27:21,480 Speaker 1: predict and project the strength of a college football team 607 00:27:21,560 --> 00:27:23,800 Speaker 1: in the year twenty twenty five. There is so much 608 00:27:23,840 --> 00:27:26,800 Speaker 1: change from last year's roster. There's so much change just 609 00:27:26,840 --> 00:27:29,320 Speaker 1: in the sport as a whole. How can we be 610 00:27:29,400 --> 00:27:32,920 Speaker 1: expected to objectively quantify what we think of this team? 611 00:27:33,200 --> 00:27:35,800 Speaker 1: That part has changed dramatically since I started doing it 612 00:27:35,800 --> 00:27:36,840 Speaker 1: publicly in twenty nineteen. 613 00:27:37,160 --> 00:27:38,840 Speaker 4: What hasn't changed, like at all? 614 00:27:38,920 --> 00:27:40,639 Speaker 1: Mike, And we'll see if the twenty twenty five season 615 00:27:40,680 --> 00:27:42,880 Speaker 1: is still to be played, But as of twenty twenty four, 616 00:27:43,000 --> 00:27:44,959 Speaker 1: as of twenty twenty three, as of twenty twenty two, 617 00:27:45,119 --> 00:27:47,560 Speaker 1: as all these other changes are happening, as roster construction 618 00:27:47,920 --> 00:27:50,639 Speaker 1: is evolving, as a transfer portal is what it is, 619 00:27:50,680 --> 00:27:54,800 Speaker 1: as NIL has come into play, what wins college football 620 00:27:54,840 --> 00:27:58,639 Speaker 1: games on Saturday has remained largely the same. Like the 621 00:27:59,080 --> 00:28:03,199 Speaker 1: elements of the game, if you do X, Y and Z, 622 00:28:03,520 --> 00:28:06,360 Speaker 1: if you do these better than your opponent you win. 623 00:28:06,760 --> 00:28:10,320 Speaker 1: So it's funny so much has changed in and around 624 00:28:10,320 --> 00:28:13,119 Speaker 1: the sport. But at the same time, what I love, 625 00:28:13,240 --> 00:28:15,119 Speaker 1: and I think what a lot of people love about 626 00:28:15,119 --> 00:28:17,639 Speaker 1: college football is Saturdays in the fall. 627 00:28:17,840 --> 00:28:19,879 Speaker 4: Like we that is what is all about. 628 00:28:19,920 --> 00:28:22,400 Speaker 1: That's why I do all this work throughout the week, 629 00:28:22,480 --> 00:28:25,040 Speaker 1: throughout the summer, throughout the spring. I do all this 630 00:28:25,160 --> 00:28:28,040 Speaker 1: work because I love sitting down on Saturday afternoon and 631 00:28:28,040 --> 00:28:30,240 Speaker 1: watching games. And at the end of the day, if 632 00:28:30,280 --> 00:28:32,840 Speaker 1: you do certain things in twenty twenty five that you 633 00:28:32,880 --> 00:28:35,360 Speaker 1: were doing in twenty nineteen and twenty fifteen and two 634 00:28:35,440 --> 00:28:37,800 Speaker 1: thousand and five, you're probably still gonna have the same 635 00:28:37,880 --> 00:28:40,240 Speaker 1: level of success. That doesn't mean the better team always wins. 636 00:28:40,280 --> 00:28:42,200 Speaker 1: You could argue that didn't happen in the Kansas State. 637 00:28:42,120 --> 00:28:42,880 Speaker 4: Iowa State game. 638 00:28:43,200 --> 00:28:45,640 Speaker 1: You know, if those two teams played again, I'd probably 639 00:28:45,640 --> 00:28:47,440 Speaker 1: still make Kansas State a favorite on a neutral field. 640 00:28:47,440 --> 00:28:49,080 Speaker 1: And that drives people nuts because they're like, they just 641 00:28:49,080 --> 00:28:50,760 Speaker 1: played on a neutral field and Iowa State one. 642 00:28:51,040 --> 00:28:52,680 Speaker 4: Yeah, I get it. I'm just telling you how Vegas 643 00:28:52,680 --> 00:28:53,120 Speaker 4: works too. 644 00:28:53,840 --> 00:28:55,680 Speaker 1: That's why they have big, shiny buildings in the desert 645 00:28:55,720 --> 00:28:59,000 Speaker 1: because they're able to capitalize on that kind of perception 646 00:28:59,080 --> 00:29:02,400 Speaker 1: in the marketplace. So it's interesting to me that what 647 00:29:02,440 --> 00:29:04,720 Speaker 1: wins college football games today is the same as kind 648 00:29:04,720 --> 00:29:07,160 Speaker 1: of what it's always been, even though everything else about 649 00:29:07,200 --> 00:29:09,000 Speaker 1: the sport seems to have seemed so changed. 650 00:29:09,960 --> 00:29:11,920 Speaker 2: That'll do it for today's episode of the Big Bets 651 00:29:11,920 --> 00:29:14,520 Speaker 2: on Campus podcast. Thanks for tuning in, and don't forget 652 00:29:14,600 --> 00:29:17,280 Speaker 2: to download the free award winning Action Network app and 653 00:29:17,360 --> 00:29:19,760 Speaker 2: leave us a five star rating and review wherever you 654 00:29:19,800 --> 00:29:22,800 Speaker 2: listen to the podcast. If you haven't already, please subscribe 655 00:29:22,800 --> 00:29:25,240 Speaker 2: to our YouTube channel search for BBOC and you'll have 656 00:29:25,280 --> 00:29:27,760 Speaker 2: access to all our preseason content, all of our Week 657 00:29:27,800 --> 00:29:31,000 Speaker 2: one plays, and all throughout the season. Every single time 658 00:29:31,000 --> 00:29:33,080 Speaker 2: that we drop an episode is going to be on YouTube. 659 00:29:33,120 --> 00:29:35,760 Speaker 2: If you'd like to follow Kelly's rankings, check out the 660 00:29:35,840 --> 00:29:39,120 Speaker 2: kforard Ratings dot com or follow them at kforward Ratings 661 00:29:39,200 --> 00:29:41,440 Speaker 2: over at Twitter. He also wrote the forward to my 662 00:29:41,480 --> 00:29:44,480 Speaker 2: book Legendary Bowl Deciding College Football's Goat. You'd find that 663 00:29:44,560 --> 00:29:47,280 Speaker 2: over on Amazon. For Kelly Ford, I'm my caliberys. Thanks 664 00:29:47,280 --> 00:29:49,520 Speaker 2: for listening to the Big Bets on Campus podcasts, and 665 00:29:49,600 --> 00:29:51,760 Speaker 2: as always, best of luck with all your bets. 666 00:30:00,160 --> 00:30:04,680 Speaker 3: Action Network reminds you please gamble responsibly. If you or 667 00:30:04,720 --> 00:30:07,480 Speaker 3: someone you care about has a gambling problem, help is 668 00:30:07,520 --> 00:30:10,600 Speaker 3: available twenty four seven at one eight hundred gambler