Analyzing ITT results for team performance problematic

BY CHARLES SCUDDER | @cscudder

Historically, the team with the winning rider of the men’s ITTs has a 75 percent chance of hoisting the Borg-Warner Trophy, while 33.3 percent of women’s ITT winners win the race.

But without analyzing the performance of every single cyclist for the past 20 years of ITT results — that’s at least 5,000 riders, for those of you keeping score at home — it is nearly impossible to produce a model for predicting how ITT performance weighs into the probability each team has to win the race.

If we rank each rider from Saturday and award points per position — first gets one point, second gets two points, third gets three, etc. — we start to get a picture of team performance in an individualized event.

MEN
Northern Indiana Cycling — 66
Alpha Epsilon Pi — 89
Beta Theta Pi — 106
Cutters — 111
Evans Scholars — 111
Phi Kappa Tau — 128
Sigma Chi — 135
Phi Gamma Delta — 156
Dodd’s House — 184
Delta Tau Delta — 221
Pi Kappa Alpha — 237
Black Key Bulls — 250
Forest — 278
Wright Cycling — 286
Alpha Tau Omega — 298
Delta Chi — 309
Phi Kappa Sigma — 350
Delta Sigma Pi — 351
Phi Kappa Psi — 353
Sigma Phi Epsilon — 360
Phi Sigma Kappa — 360
Gray Goat — 403
Sigma Nu — 409
Kappa Delta Rho — 481
Sigma Alpha Mu — 495
Sigma Alpha Epsilon — 547
Delta Upsilon — 574
Phi Delta Theta — 590
CSF — 600
Lambda Chi Alpha — 697
Collins — 965
Sigma Pi — ND
Pi Kappa Phi — ND

WOMEN
Army — 113
Mezcla — 131
Chi Omega — 162
Cru — 163
Delta Sigma Pi — 168
Ski — 181
Kappa Alpha Theta — 185
Melanzana — 193
Alpha Sigma Alpha — 227
Theta Phi Alpha — 240
CSF — 246
Alpha Chi Omega — 255
Kappa Kappa Gamma — 262
Alpha Xi Delta — 268
Teter — 284
Sigma Delta Tau — 288
Phi Mu — 316
Zeta Tau Alpha — 316
Delta Zeta — 345
Delta Gamma —358
Alpha Omicron Pi —360
Alpha Gamma Delta — 389
Wing It — 392
RideOn — 400
Gamma Phi Beta — 411
Alpha Delta Pi — 417
IU Nursing — 460
Delta Phi Epsilon — 497
Collins — 650
Kappa Delta — 782
Air Force — ND
Delta Delta Delta — ND
Alpha Phi — ND

But those statistics aren’t exactly fair, either. Northern Indiana Cycling, which scored 66 points in our model, only had one rider who placed 66th. Collins, which placed dead last in the men’s rankings, had eight riders, making their point total much higher.

Some quick averaging of team totals over the number of competing riders in each team shows the average ITT placement for each qualified team.

MEN
Beta Theta Pi — 26.5
Cutters — 27.8
Black Key Bulls — 27.8
Delta Tau Delta — 36.8
Alpha Epsilon Pi — 44.5
Sigma Chi — 45
Wright Cycling — 47.7
Sigma Phi Epsilon — 51.4
Phi Gamma Delta — 52
Phi Delta Theta — 53.6
Sigma Alpha Epsilon — 60.8
Northern Indiana Cycling — 66
Forest — 69.5
Phi Kappa Sigma — 70
Delta Sigma Pi — 70.2
Gray Goat — 80.6
CSF — 85.7
Lambda Chi Alpha — 87.1
Phi Kappa Psi — 88.3
Dodd’s House — 92
Delta Upsilon — 95.7
Alpha Tau Omega — 99.3
Sigma Nu — 102.3
Delta Chi — 103
Evans Scholars — 111
Pi Kappa Alpha — 118.5
Phi Sigma Kappa — 120
Kappa Delta Rho — 120.3
Collins — 120.6
Sigma Alpha Mu — 123.8
Phi Kappa Tau — 128
Sigma Pi — ND
Pi Kappa Phi — ND

WOMEN
Cru — 32.6
Army — 37.7
Melanzana — 38.6
Alpha Chi Omega — 42.5
Mezcla — 43.7
Ski — 45.3
Kappa Alpha Theta — 46.3
Teter — 47.3
CSF — 49.2
Phi Mu — 52.7
Alpha Xi Delta — 53.6
Chi Omega — 54
Theta Phi Alpha — 60
Kappa Delta — 60.7
Alpha Gamma Delta — 64.8
Wing It — 65.3
Kappa Kappa Gamma — 65.5
Delta Gamma — 71.6
Alpha Sigma Alpha — 75.7
Collins — 81.3
Gamma Phi Beta — 82.2
Delta Sigma Pi — 84
Alpha Omicron Pi — 90
Sigma Delta Tau — 96
Ride On — 100
Alpha Delta Pi — 104.3
Zeta Tau Alpha — 105.7
Delta Zeta — 115
IU Nursing — 115
Delta Phi Epsilon — 124.3
Air Force — ND
Delta Delta Delta — ND
Alpha Phi — ND

That looks a little more accurate, doesn’t it?

With Miss ‘N Out and Team Pursuit delayed until next weekend, it’ll be longer until we get a better picture of team probability of winning the race. But expect standout performances from ITT champs Aryn Doll of Chi Omega and Christopher Craig of Beta Theta Pi.

UPDATE: Shortly after posting the above numbers, we received some thoughtful feedback from a handful of our readers, pointing out something we hadn’t thought about.

Truthfully, a handful of rookies and other riders that will not compete on race day compete in ITTs. In our original post, we included every single one of those riders as well, showing each team’s overall performance in ITTs.

In order to show a better picture of what each team will look like on race day, we’ve taken only the placement of the top four riders for a given team. We then averaged that total by the number of ITT riders per team, maxing at four.

MEN
Phi Delta Theta — 7.8
Black Key Bulls — 9
Delta Tau Delta — 24.8
Beta Theta Pi — 26.5
Cutters — 27.8
Sigma Phi Epsilon — 31.8
Phi Kappa Tau — 32
Sigma Alpha Epsilon — 35
Wright Cycling — 39.5
Alpha Epsilon Pi — 44.5
Sigma Chi — 45
Phi Gamma Delta — 52
Phi Kappa Sigma — 58.6
Lambda Chi Alpha — 61.5
Delta Sigma Pi — 62
CSF — 63.3
Northern Indiana Cycling — 66
Forest — 69.5
Gray Goat — 70
Delta Upsilon — 80.3
Phi Kappa Psi — 88.3
Dodd’s House — 92
Alpha Tau Omega — 99.3
Sigma Nu — 102.3
Delta Chi — 103
Collins — 106.3
Evans Scholars — 111
Pi Kappa Alpha — 118.5
Phi Sigma Kappa — 120
Kappa Delta Rho — 120.3
Sigma Alpha Nu — 123.8
Sigma Pi — ND
Pi Kappa Phi — ND

WOMEN
Teter — 10.5
Kappa Alpha Theta — 16.5
Alpha Chi Omega — 18
Melanzana — 26.8
Phi Mu — 27.8
Cru — 29.5
Army — 37.7
Alpha Xi Delta — 41.5
Mezcla — 43.7
Wing It — 44.8
Ski  — 45.3
CSF — 46
Kappa Delta — 49.8
Chi Omega — 54
Alpha Gamma Delta — 58.3
Theta Phi Alpha — 60
Collins — 61.3
Delta Gamma — 63.3
Kappa Kappa Gamma — 65.5
Alpha Sigma Alpha — 75.7
Gamma Phi Beta — 76.8
Delta Sigma Pi — 84
Alpha Omicron Pi — 90
Sigma Delta Tau — 96
Ride On — 100
Alpha Delta Pi — 104.3
Zeta Tau Alpha — 105.3
Delta Zeta — 115
IU Nursing— 115
Delta Phi Epsilon — 124.3
Air Force — ND
Delta Delta Delta — ND
Alpha Phi — ND

What I find most interesting about these new adjusted averages is the top of the men’s field, where there is a 1.2-point difference between first and second position, but a 15.8-point difference between second and third. We can expect a strong showing from both Black Key Bulls and Phi Delta Theta on race day. The top riders on each of those teams are markedly stronger than those of other teams in the field, according to their ITT performance.

Charles Scudder is a senior studying journalism at IU. He put off taking a required statistics course until the last semester of his senior year, so go easy on him if his math is screwy. Email him at cscudder@indiana.edu.

16 thoughts on “Analyzing ITT results for team performance problematic

  1. I would make one more adjustment..
    only factor the top 4 riders on each team.
    this model adds rookie riders’ times that wouldn’t show up on race day

    Love the site and great work!

  2. Bikes and Math is correct. Some additional ways to play with the data would be to average actual times instead of places …… or to examine which team has the strongest 2 through 4 riders……etc. Thanks for the articles.

  3. I did some similar math, and it’s interesting to note that taking the average times of each teams best 4 riders produces slightly different standings than this methodology!

  4. Because so many times are separated by a a mere few tenths of a second and some teams have more than 5 riders “scoring” points using a ranking system you do end up with a different and perhaps clearer picture if you average each teams top 4 riders’ times.

    I like to break down the results by 1 second intervals and feel that a 4 seconds difference represents a significant difference (1 second per lap)….so, if you round to the tenth and look at the top 30 riders……:
    2:23s – 4 Riders
    2:24s – 3 Riders
    2:25s – 5 Riders
    2:26s – 6 Riders
    2:27s – 12 Riders

  5. They don’t make you take a stats class in journalism school do they? These facts and figures simply aren’t predictive of anything

    • Go easy on him, he’s putting off his stats class until senior year!
      If you scroll to the bottom, there’s some corrections too

  6. You defiantly want to do average times instead of average place. Average place is a misleading category. so if there is .03 between the 25th rider and 30th rider one person get’s 5 extra points.

    Really need to think this data stuff out Charlie.

    Average times….

    If you want to predict winners, you need to find times of top 3 riders. The 4th rider is extremely overrated.

    • I agree that average time is a better way to analyze the results than placing.

      The forth rider is “extremely overrated” only if you don’t ride for Phi Delt or BKB, or unless you plan on having him not ride in the race. Otherwise, the last time I checked, you are allowed to ride up to four people on race day.

      • 4th rider is extremely overrated, since I have already run my own analysis, and there is no correlation between 4th rider performance and your place. The 4th rider ca be a top 100 rider, anywhere in there.

        While all the top 4 teams over the last 3 years have all had there top 3 riders finsh top 35 in ITT. The 4th rider is extremely overrated. all he need to do is staff with the peleton. That’s all, and that can be considered very easy.

  7. DATAANALYTICS–
    Hopefully, your data analysis is better than your grammar because your reply is difficult to follow.

    Let’s see your analysis. Post it.

    Three years worth of data points can hardly be considered statistically valid. I don’t disagree that if you have an Eric Young or a Hans Arneson that derivates himself so much from the rest of the field that he can offset the lack of fitness from the 2-4 riders, but none of the riders in this year’s field is that good. Therefore, unless your roster is as deep as Phi Delt or BKB, as soon as you put on your weak 3/4 guys on the bike, the team across the board has to waste matches because of lack of fitness or compensation for weaker teammates.

    • 1. My anlaysis is over last 10 years. Not 3, I miss typed.
      2. It’s a message board not a English paper, so do not care about grammar.

      Analysis is in excel and is many pages deep. So it is not easy to just post it. I am not wasting an hour trying to figure out how to post it. Do your own work.

      Plus my analysis is done for one of the cycling teams. Not to be made public. This team has done VERY well last 5 years. So… my analysis speaks for itself.

      Just let me tell you, the 4th rider is overrated. Bottom line. Last 12-15 laps are done by 1-3 riders. TYPICALLY. The 4th rider needs to be able to ride a bike, yes, but overrated.

      Your best at actually matching the times of the top 3 riders. That will be even better indication.

      • @Data,

        I agree with your point that typically the 4th rider does not need to contribute as much as 1-3 and I respect that you don’t have to share your information with everyone.

        But I believe a 4th rider can play a role in the last 12-15 laps and is vital for the team to get to that point.

        Look at Beta’s strategy from last year, that doesn’t work unless every rider does a set.

        Also, while you don’t need a 4th rider to win, you need one to not lose.

        You said that typically the 4th rider needs to ride a bike or get a top 100 ITT, but having that rider out puts the team at risk of getting dropped if other teams’ big guns are in at the same time. You can still pull a top 5-10, but you won’t win

        Even if all that is asked is to stay in the pack and prevent escapes, the 4th rider can eat laps and get the 1-3 riders some rest that will pay off later in the race.

        Congrats on your teams success and thanks for sharing

  8. Bikes and Math,

    Thanks for your response. Yes, I agree with your points about 4th rider. I certainly think he plays a role and can be an excellent part of your team. No doubt about it. Just saying there is no correlation between 4th riders ITT time finish. The 4th rider can finish pretty much anywhere. It’s the top 3 riders that have correlation to team finish. Obviously a good 4th rider is only a good thing. Just not a requirement.

    The team with best 4 riders doesn’t typically finish better than the team with the better top 3 riders. That is all.

  9. Pingback: Team Pursuit gives better idea of race winner, final probability | 33to1

Join the Discussion

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s