U. Dayton BasketballKenPom Has Evil Urges

kenpom16Kenneth Pomeroy’s annual pre-season projections are like the sun — their existence confuses Native Americans (both are the white man’s voodoo) and they are fun to look at for an hour or so. 

I am fascinated by the prognostication racket, the fact that the people who are its biggest detractors are almost without exception its most ardent followers. The prediction business has transcended sports, most notably spilling over into the stock market and even political races. Projections are frivolous, often meaningless, and people can’t get enough.  

Pomeroy is the leading college basketball fortune teller. I pay for his website every season — so I too am a rube, lined up at the statistical ring toss booth eager to give my money away.

A look at Ken Pom’s forecast indicates that he is not quite as bullish on the Flyers as other outlets seem to be.

Kenny predicts the Flyers will finish around 18-10 on the season, 11-7 in Atlantic Ten play. This estimate does not include the two games UD will play after facing Iowa in the Advocare Invitational in late November. (If you actually add up the wins and losses, it seems Pomeroy has Gem City winning 22 games and losing 6, yet it clearly indicates 18-10 as the final projection. Although this seems like a glaring error, I’m sure there is some analytical explanation for this inconsistency. Leave your interpretation in the comments if you have an idea why the numbers don’t match up.)

As far as the conference race is concerned, KenPom projects Davidson and Rhode Island to finish tied for first with a 12-6 record, while VCU, George Washington, Richmond, St. Joe’s (??) and Dayton are all in a logjam for third with 11-7 records. I am by no means an “advanced analytics” Luddite, but if St. Joseph’s finishes 11-7 in the A10 this season I will watch every episode of Katie Nolan’s Garbage Time in one sitting.

Remember, these projections have no meaning unless you get riled up about them. You are better off viewing them in the same manner you read this post, with a slight shrug before moving on to barely more important things.

Tom Blackburn

Tom Blackburn is a proud U. o' D. alum. He loses faith in humanity one day at a time, but not in you, you seem like you are all kinds of alright. Charter member of the T-Man fanclub.


  • Big 10

    October 27, 2015 at 9:18 AM

    I believe he was wrong on us the last 2 years


  • tribebrowns

    October 27, 2015 at 9:39 AM

    His projection of wins and losses is based on the probabilities of win / loss not each actual W / L he puts up there. If you have alot of 52% win probabilities your projected record will deviate much lower than if you just count Ws.


    • Tom Blackburn

      October 27, 2015 at 9:54 AM

      Thank you, sir.


  • tribebrowns

    October 27, 2015 at 9:47 AM

    NERD ALERT: Having studied and even tracked Pom’s projections for a lengthy amount of time, they should be taken with a grain of salt. He’s no better than flipping a coin when it comes to how his projections perform against the spread. The guy has good general projections, but is by no means all that precise. He likely had a difficult time including Cook’s contribution and may have left Dyshawn out. He also typically only quantifies contribution of freshman if they are top recruits. (This info about his approach is somewhere in his blog archives I forget).

    The most useful part of his site is a data source for advanced (pace-neutral) and individual player stats. His projections are probably slightly better than AP yahoos, but not substantially all that better.


    • Dyshawn's Fluffer

      October 27, 2015 at 11:28 AM

      Don’t think you needed the nerd alert disclaimer with a moniker like Tribebrowns


  • The Morgan Miller Fathead above Driesen's bed

    October 27, 2015 at 10:12 AM

    I’m wearing a diaper


  • Edwin C Moses

    October 27, 2015 at 1:24 PM

    There is no way we win the conference or have enough to go dancing without #21.


    • Tom Blackburn

      October 27, 2015 at 1:37 PM

      I partially agree, wanna see what we get out of Cooke before I go all in.


      • Sportz Fan

        October 27, 2015 at 2:48 PM

        I hope Cooke is a better player than he was a witness.


        • Tom Blackburn

          October 27, 2015 at 2:51 PM

          Damn that’s good.


        • TBolt

          October 28, 2015 at 1:45 PM

          Unless Cooke was covering for his teammate, which is very plausible.


  • Angry John

    October 27, 2015 at 1:27 PM

    That is some funny math.


    • Humbert

      October 27, 2015 at 9:53 PM

      Actually, it isn’t. Tribebrowns explained it above. It takes all of a second grade education to understand.


    • TonyC

      October 28, 2015 at 3:20 PM

      Spoken like someone owning a PridePlus membership.


    • John Deere

      October 30, 2015 at 10:25 PM

      Angry John, a little off topic, but you’ll be glad to know that all UD games will be webcast in Mandarin this year.


  • udflyerfan07

    October 27, 2015 at 9:24 PM

    At least he gives us a chance to go undefeated this year.


  • Fletch

    October 27, 2015 at 10:25 PM

    His projections are useless until after Christmas. They are pretty good after 12-15 games worth of data.


  • TBolt

    October 28, 2015 at 9:58 AM

    Isn’t Ken Pomeroy the guythat gets Swampy’s shit stained Fruit of the Looms all in a bunch?


  • Matt CT

    October 28, 2015 at 3:16 PM

    The reason Ken Pom’s projected record comes out to 18-10 is due to the effects of the combinatorial probability effect of each game. To demonstrate, let’s take the subset of the 10 games where Dayton is favored to win at least 80 percent of the time. Dayton can expect to win all of those games just 17.09 percent of the time….and in all likelikhood will lose 2 of those games. For the other games where Dayton is projected to win by a lower percentage, several of those will end up as losses as well. For the games where Dayton is projected to lose the Flyers can be expected to lose all six just 13 percent of the time, indicating they’ll likely steal a win back there.

    Certainly if every win and every lose was listed at 100 percent, 22-6 would be the projected record.


  • Nostradamus

    October 28, 2015 at 5:27 PM

    I heard a rumor they planned to actually play the games to see what happens.


    • Swampy

      October 31, 2015 at 4:31 PM

      Lose to Iowa = no Dance. 2 terrible rpi teams after that. Lose to Vandy = no Dance. Lousy non con schedule. Wonder why UD was picked last in 2015 Dance? This is why. They don’t play anybody. A-10 has 103 avg rpi. Terrible. Need to schedule tough teams and beat them in non-con. I’m talking “Nova, Duke, Wichita St. , etc. Top 25 teams.


      • TBolt

        November 2, 2015 at 12:53 PM

        Did you ever get a response from Coach K?


  • Swampy

    October 31, 2015 at 4:44 PM

    114 avg rpi of above w/o Orlando loser bracket games.


    • Sweaty Balls

      November 1, 2015 at 7:31 PM

      I know you are an imposter, you didn’t post any links


      • The kid from Lowes

        November 2, 2015 at 1:13 AM

        Nice move


  • Gil

    November 3, 2015 at 2:21 PM

    I jerk it to Katie Nolan’s Shitty web series on the reg


Leave a Reply

Your email address will not be published. Required fields are marked *

Related Posts