Monthly Archives: March 2012

A proposal for NFL overtime

Earlier this week, I linked to an interesting auction-based proposal to help improve fairness in NFL overtime games. Right now, the coin flip gives the winning team a boost more often than not (the only exception is if the winning team mistakenly takes the ball but has a VERY weak offense relative to its defense, or, likewise, if the losing team has a VERY strong defense, relative to its offense).

The idea of the auction is to give each team “accurate” odds of winning by having them bid for the ball, using starting field position as currency. As you bid to start deeper and deeper in your own end, the odds of you scoring before your opponent drop. At some starting field position for each team (maybe around the 17 yard line), the odds should be close to fair Continue reading

Third/fourth round results and final four simulations

Unfortunately, I didn’t move on to the final four of the TeamRankings blogging competition. It was fun while it lasted though, and thanks again to TeamRankings for putting it on and providing great data (which I will continue to use for NCAA tourney simulations).

It was a great third round and a so-so fourth round for my simulations. Here’s the update on the initial brackets that I’ve been tracking:

I set up the first four brackets by always choosing the “better” team according to the ranking listed on the left. The last row uses my simulations to pick the winner. I stumbled a little in round 2, but recovered strongly in the elite eight (6 out of my initial 8 predictions made it, with only Missouri and Michigan State coming up short). Pomeroy and Sagarin’s rankings proved the best at predicting the final four — both missed only Louisville (they each had Michigan State). I missed Louisville and Ohio State (I had Syracuse, by a nose).

If I forgive bracket mistakes and re-pick each game based on who actually played, here are the success rates Continue reading

Round 2 post for TeamRankings competition

My post for round 2 of TeamRankings blogging competition is up on their blog. I give a full explanation of my NCAA basketball simulation method and flesh out my predictions for tonight’s games.

Edit: The complete post can now be found below. I moved it here in case TeamRankings changes their links at a future date.

Breaking Down Match Ups: Sweet Sixteen Game Simulations

In round 1 of the Stat Geek Idol competition, I described a procedure to simulate NCAA basketball games based on the few team statistics that really matter: shooting percentages, shot selection, turnovers per play, and offensive rebound percentage. These are basically Dean Oliver’s four factors, though I go a little more in depth. For this round, I’ll break down the simulation procedure and apply it to the Sweet Sixteen match ups. But first, how have my simulations performed so far? For comparison, I list the number of teams correctly predicted to reach the second and third rounds by a few different methods (I give a full summary on my blog):

  • Take the higher seed: 22/32, 11/16
  • Take the higher RPI: 21/32, 9/16
  • Take the higher Pomeroy ranking: 22/32, 10/16
  • Take the higher Sagarin ranking: 23/32, 10/16
  • Take the team that wins majority of my simulations: 23/32, 9/16

If I forgive first round mistakes and recalculate second round match ups Continue reading

Simulated stats for the sweet sixteen

Over the past few posts, I’ve been focusing on the NCAA tournament, simulating games based on predicted efficiency statistics. For the Sweet Sixteen predictions below, I ran 8,000 simulations for each game. I list my predicted winner (including 7 Florida over 3 Marquette), and the predicted efficiency statistics. The stats are based on Dean Oliver’s four factors:

  • Factor 1: 3 pt shooting %, 2 pt shooting %, foul shooting %
  • Factor 2: % of potential offensive rebs secured (including balls out of bounds)
  • Factor 3: % of offensive plays ending in a turnover
  • Factor 4:  3 pt attempts as a % of non-turnover plays, 2 pt attempts as a % of non-TO plays, free throw trips as a % of non-TO plays

Factor 4 is the most confusing. It’s similar to Oliver’s FTA/FGA factor, but has more value for simulations, since it tells me how often teams get a three point attempt, a two point attempt, or a trip to the line (on plays without a turnover).

1 Kentucky, 4 Indiana Favorite: Kentucky (wins 55.3% of simulations):

  • 2 pt %: 50, 46
  • 3 pt %: 35, 39
  • FT %: 72, 76
  • OReb %: 34, 30
  • TO %: 14, 14
  • 2 att %: 62, 64
  • 3 att %: 23, 22
  • FT att %: 15, 14

3 Baylor, 10 Xavier Favorite: Baylor (76.9%): Continue reading

Simulation results through NCAA tournament round 2

On Saturday, I posted the round 1 performance of my NCAA tournament simulations, using data from Teamrankings.com. I did pretty well: 23/32 games correct, similar to some other prediction methods that I tested. Before round 3 kicks off, I wanted to go through my results through round 2. For comparison, in the first four rows of the table below, I took the better team in each game as indicated by the ranking listed on the left (i.e., the higher seed, the team with the better RPI, etc.). For the “Causal Sports Fan” row, I took the team that won the majority of my game simulations.

My picks to make the sweet sixteen did a little bit worse than my first round picks — only 9 out of 16 are still in. However, some sweet sixteen picks dropped out in the first round for each method (I lost 2 Missouri, 2 Duke, 4 Michigan, and 11 Texas). If I forgive first round mistakes and examine the actual round two match ups, all methods did quite well in the second round: Continue reading

NCAA tournament simulations post at Teamrankings.com

My first round submission in the Stat Geek Idol competition is up at the TeamRankings blog today. Check it out, along with the other contest entries. Second round submission coming soon . . .

If you’d like to read more about my tournament predictions, check out my recent post on round 1.

Optimal drafting

There’s a new paper out this month by Casey Ichniowski and Anne Preston concerning the NCAA tournament and the NBA draft (thanks to my PhD buddies Chris and Felipe for passing it along). Their argument is that unexpectedly strong tournament performance (especially team performance) causes players to be selected earlier in the NBA draft. This isn’t a bad thing though — in fact, they suggest that these strong tournament players tend to outperform their draft position in the NBA.

I believe their results saying that tournament performance affects draft position (this has also been shown by Chaz Thomas in an undergraduate thesis, and by David Berri, Stacey Brook, and Aju Fenn), and I mostly believe their results that strong tourney performers should be drafted even earlier, though their set up is a little odd for this second issue.

The clearest way to show that teams make mistakes in the draft Continue reading

NCAAs round 1 results

For the NCAA tournament this year, I simulated each game based on my predicted efficiency stats for each team (shooting percentage, shot selection, turnover rate, and offensive rebound rate). I submitted my work for Teamrankings.com’s college basketball blogging competition and I’m thrilled to announce that I moved on to the next round! I’m pretty excited for round 2 (deadline Tuesday at midnight). If you have a topic suggestion, let me know — bonus points for something semi-related to the game simulations I’ve been running.

Now how about those simulations? Here’s a performance overview against a few other rankings (seeding, RPI, Ken Pomeroy, and Jeff Sagarin):

Simply choosing the higher seed got 22 games right. RPI did slightly worse at 21, and I matched Sagarin at the top with 23 correct (71.9%). The problem? As evidenced by the potential wins columns, I lost my champion. Missouri battled Norfolk State the whole game and came out behind. Every method lost Missouri, Duke, and Michigan as round 2 predicted winners, and Missouri, Duke, or both as round 3 winners. I’ll need some help to catch up at the end, though. No other system had Missouri or Duke advance to the Final Four.

Specifically, where did I go right? Of my “upset locks” (over 60% probability), VCU and NC State came through, and Alabama nearly beat Creighton. It’s worth reminding at this point  Continue reading

Byebye, Nene

When the Nuggets resigned Nene to a large contract in the off season, I wrote that they were making the most of a bad situation (few marquee free agents available, Nene not worth the money). Today, they admitted as much by trading Nene in a three team deal that landed them talented but unreliable JaVale McGee.

Although the Nuggets have bucked the trend slightly by having some success post-Iverson and Carmelo, my preferred models for NBA success are the South Beach Talents and the Seattle Scientists. The Talents involve trading or signing two (or three!) top 15 players, and spending the spare change on spare parts to fill out the roster. The theory is that mid range guys are overpaid, so just spend you money at the top and bottom.

If you’re a small market, it’s tough to attract stars these days, so I suggest the Scientists, a hypothetical NBA team that attacks a different undervalued asset: effort. The Scientists hire guys who will be in better shape and work harder than every other team. They press on offense AND defense.

The Nuggets are neither of these. They are doing okay with their strategy of paying mid range guys (Afflalo is another example); they have the 6 seed right now. However, they probably aren’t a championship team. Wouldn’t it be fun to try something new instead?

Adventures in picking the NCAA tournament, part 2

Thanks to Teamrankings for the data for this work. I hope that my models someday forecast as well as yours!

Yesterday, I tried some approaches to predict the NCAA tournament. My favorite of these is where I use team efficiency stats (shot selection, shooting percentage, turnovers per possession, and offensive rebounding percentage) to simulate whole games possession by possession. This approach predicts win-loss correctly in about 70% of tournament games over the past five years, so I decided to predict the whole tournament with this method for this year.

Below are the win probabilities that I generated for each region (the percentages on each line give the probability of winning the previous game).

The South:

Kentucky is the odds on favorite to win it all, and I predict that they will get out of their region. Their stiffest challenge could be Indiana, Continue reading