G’s Explanation or Just a Couple of Little Things

Happy Wednesday college football fans. We had an eventful extended week 1 and I can’t wait to see what surprises are in store for us this weekend. For those that are new or newer to the GCR, we have 4 different kinds of posts. First is G’s Evaluation which is the computer ranking based on games to date – more on that in a moment. Second is G’s Explanation which is a bit of analysis into either why the rankings are what they are (1/2 of today’s post) and/or interesting (at least to me) statistical data about the season so far (the other 1/2 of today’s post). Third is G’s Exploration which is used mostly in the off-season, but occasionally during the season where we explore things like the preseason SOS and/or conference comparisons. Finally, we have G’s Expectation which is used to list and predict games as well as highlight fun or meaningful games to watch, always produced the day before the first game of the week or during the first game of the week.

…..

If you looked at Tuesday’s G’s Evaluation (by the way, go to http://bit.ly/2019GCR to register your email to get instant posts or to review past posts – always free), you may be confused at the rankings. For example, the consensus #1 team in the polls, Georgia is 35th. Why the discrepancy? First, the GCR has no preseason ranking: everyone started out tied for first. Teams that play down (e.g., Power 5 Georgia playing FCS Tennessee Martin) get a…discount on the win. Over the course of the season and the fact that all but about a dozen Power 5 teams play an FCS school at some point, the discount is washed out, but when the denominator is 1, it matters. Future games may not be played, so strength of schedule at this point for most teams is they lost to a 1-0 team or they beat an 0-1 team. It doesn’t matter that Duke beat Clemson (who was a top 10 team in the polls), the GCR just sees that they beat a Power 5 team with a winning percentage of .000. So, the GCR is a bit of a puzzle during the first couple of weeks. The cream will rise and the whey will sink (if the whey actually sinks, who knows?) as the season progresses. The other thing to consider is that, if Clemson wins out from here, Duke will get credit for beating an 11-1 team. The GCR considers all games that have been played to be happening…now.

…..

All of that to say, who would have thought LSU, TCU, Baylor, and Clemson would all have losses after one game? It’s going to be a great, sometimes upsetting, sometimes exhilarating, sometimes wacko season.

…..

Each year, I try to come up with something new. A different statistic than I’ve used before and that I’ve not seen anywhere else. This year, I have a quality of game measure. I’ve talked to a few people who are a little fed up with the mismatches when teams play outside of their league (Power 5, Group of 5, FCS). So a “quality game” could be defined as a game between two teams in the same league. I know two 5-0 teams playing is a better game than a 5-0 team vs an 0-5 team, but I’m trying to keep this somewhat simple to see how it works. There are 6 “types” of games that can occur: Power 5 vs Power 5, Power 5 vs Group of 5, Power 5 vs FCS, Group of 5 vs Group of 5, Group of 5 vs FCS, FCS vs FCS, and FCS vs non-Division I. I gave each type a score. Like leagues are worth 3 points each, even if two 0-5 teams play each other. A P5 team vs a G5 team got a value of 1 because this is the most common “league-upset” combination. P5 vs FCS loses 3 points, G5 vs FCS loses 2 points, and FCS playing out of league loses 1 point. Here are the extended week 1 numbers.

TypeFactorNumber of GamesPointsCumulative GamesCumulative PointsAverage
P5 vs P53113311333.00
G5 vs G53113322663.00
FCS vs FCS334102561683.00
P5 vs G513030861982.30
FCS vs Non D1(1)16(16)1021821.78
G5 vs FCS(2)23(46)1251361.09
P5 vs FCS(3)19(57)144790.55

…..

So, with a maximum of 3.00, the mismatches brought the overall average to 0.55. Only 56 of the 144 games (.389) were within their own league. I don’t think I’m alone that the fans feel this statistic. There were a few marquee games, but many of them were designed to get a quick win (although that didn’t always happen) or get a quick paycheck to fund the rest of the season. If I was put in charge of NCAA football, I’d have a preseason game or two, that wouldn’t count, that was designed to let teams get game experience against weaker or much stronger foes prior to the start of the season. Rules could be relaxed, etc., to try to avoid injuries. But I digress. We all hope everyone got those meh games out of the way, but no. Here is the table for this week for comparison.

TypeFactorNumber of GamesPointsCumulative GamesCumulative PointsAverage
P5 vs P53164816483.00
G5 vs G53123628843.00
FCS vs FCS32987571713.00
P5 vs G511818751892.52
FCS vs Non D1(1)13(13)881762.00
G5 vs FCS(2)20(40)1081361.26
P5 vs FCS(3)18(54)126820.65

…..

So, maybe not so much. 57 of 126 games (.452) are intra-league so we are moving up a little, but the overall average moved from 0.55 to 0.65. We will need more data to categorize, but I’m going to go out on a limb and say any value less than 1 indicates an over-arching non-competitive slate. That’s not to say there aren’t some good games this weekend. There are for sure. There are also good games that are not “in-league”. Tune in tomorrow for a listing of all of the games with as many predictions as possible.

…..

That’s it for today. Thanks for reading and sharing with others (don’t forget to share the bit.ly GCR2019 so that we can continue to grow our readership. G