Hello fellow Blather-addicts, welcome back to another edition of Pitt by Numbers; the tempo-free statistical post that gets by on blue-collar toughness and physicality.
Throughout the pre-conference slate we were told that this is Jamie Dixon’s deepest team yet. The early repetition of this proclamation has led a few of the regulars on this blog to go the other way and call our bench overrated. Now, in the wake of Pitt’s 3-0 performance in Ashton Gibbs’ absence we’re back to lauding what guys like Trevon Woodall and Talib Zanna bring to the table.
We can spend all day throwing out qualitative statements about Pitt’s depth, but just how deep are the Panthers? More importantly, how much does depth matter?
I wanted to see if there was some kind of trait that teams who make deep runs into March have in common when it comes to how they rely on their bench. I took the last four years worth of Final Four teams and compared the percentage of minutes, points, rebounds and assists that came from their benches. Here is how they compare to this year’s Pitt team. (You may have to maximize your browser to view the graph)
| Bench Min. | Points | Rebounds | Assists | ||||
| 2010 | Duke | 23.4% | 16.8% | 32.4% | 12.4% | ||
| WVU | 25.8% | 21.8% | 30.7% | 31.9% | |||
| Butler | 23.3% | 18.1% | 26.2% | 18.3% | |||
| MSU | 32.4% | 25.7% | 40.7% | 32.3% | |||
| 2009 | UNC | 27.4% | 18.4% | 39.9% | 27.7% | ||
| MSU | 36.4% | 35.3% | 49.6% | 30.6% | |||
| Nova | 29.0% | 23.8% | 44.8% | 21.1% | |||
| UConn | 25.5% | 26.5% | 33.9% | 29.4% | |||
| 2008 | Kansas | 29.5% | 30.1% | 38.3% | 32.2% | ||
| Memphis | 32.4% | 29.2% | 31.7% | 28.6% | |||
| UNC | 21.9% | 25.5% | 33.5% | 62.7% | |||
| UCLA | 30.5% | 15.6% | 31.9% | 11.0% | |||
| 2007 | Florida | 27.6% | 24.4% | 32.4% | 22.1% | ||
| UCLA | 26.5% | 20.6% | 35.1% | 15.6% | |||
| G-Town | 32.1% | 17.8% | 29.9% | 19.6% | |||
| OSU* | 20.6% | 28.9% | 43.3% | 17.3% | |||
| *Note- Bench numbers were inflated by the performance of stud walk-on Mark Titus | |||||||
| Average | 27.8% | 23.7% | 35.9% | 25.8% | |||
| 2011 | Pitt | 33.6% | 30.2% | 41.1% | 32.6% |
What can we take from all this? Well, not a whole lot. Ideally I would have liked to have found +/- ratios for these guys to go along with my research but I had to make due with the data I could find. It’s not perfect and it largely ignores defense but it’s at least something.
The only real constant is that 15 of the 17 teams get a disproportionately high number of rebounds from the bench but I expect that to be typical of all teams, not just good ones. Big guys find themselves either tired or in foul trouble more often than guards so a disproportionate amount of the bench minutes are going to be allocated to big men. Also, rebounding comes down to effort and it’s easier to give a great effort with fresh legs.
Pitt seems to be on the higher end of the percentage of minutes coming off the bench spectrum but I expect that number to dwindle as the season goes along. Even if the bench minutes stay where they are, it probably isn’t a huge advantage or disadvantage.
Basketball is the one team sport where the best players participating have the biggest impact on the final result (admittedly, my lack of hockey knowledge should probably stop me from making such a statement) and in 99.99% of college basketball games, the best players aren’t coming off of the bench.
So is Pitt’s bench under or overrated? Who knows, but nobody has ever won a National Championship because they had an elite bench. As much as we make about the development of guys like Trevon Woodall and Talib Zanna, their development will have a much bigger impact on Pitt’s run in 2013 than this year’s tournament where Pitt will need Wanamaker, Gibbs and Brown to make the plays necessary to finally get to the Final Four.
One item of note. The first ever Pitt by Numbers post game meet-up will be held on Saturday at Foley’s on 33rd between 5th and 6th Ave. (across from the Empire State Building). Hopefully we will all be in a celebratory mood.
Update: I won’t be able to make it to Foley’s after the game and Chas won’t be in town but by all means, go to Foley’s anyway.
The fact that there is no strong correlation in any of these team’s numbers I think indicates that there is no one profile for how to effectively use your bench and if there big differences between these 16 then the other 60 each year are probably going to be all over the map also.
We also know getting to the tournament and winning deep in the tournament are two different things. Especially with a team like pitt that plays to those two extremes, great regular season record and underachieving in the tournament relative to their seeding. It might be interesting to compare even just those final four team’s regular season bench stats vs their tournament bench stats if possible.
Anyway, a few weeks ago I did the exact comparison Chas did and buried my head when I saw the exact same results. I too was a fan of bench strength but it really doesn’t seem to matter.
Indeed Pomeroy’s model doesn’t even give bench minutes “miscellaneous” status, meaning it doesn’t contribute anything statistically to winning games.
While I studied statistics I am still a big fan of an N of 1. Remember the MSU tourney game when we petered out?
What would be interesting to see along with the historical numbers above, is what those teams’ benches did during the regular season. Were they as dependent? Were they the same, or even less?
Many of us on this blog are approaching the tournaments with the thought that our success this season is directly related to our depth, and it may be. We also believe that depth gives us more options against other teams. But another conclusion one could draw from the above numbers, is that we’re overly dependent on our bench–that maybe our starters don’t consistently put up the same numbers as other teams’ starters.
I’m not a stats’ major by any means, pardon the pun!
Bench value, in my mind, is 70% about the bench’s ability to maintain what the starters achieve while the starters sit, and 30% ability to adapt to different scenarios. The first element I think could be measured by some value over replacement or i guess +/- that Pabs mentioned. The don’t know if/how the other element is measurable.
I would be curious to see, of those final four teams above, 1) how many turnovers are attributable to the bench; 2) what’s the shooting % of guys ocming off the bench.
You can conclude Pitt this year exceeds the averages of those teams, but we don’t know if the averages themselves indicated anything above normal.
Didn’t know Nate Silver posted on this blog. Thanks for collecting those stats.