Lacrosse Analytics

[jimmywork]

Your model has Penn v Army and UVA v RMU both as coin flips. We'll see how those play out, but it sure seems like your ELO model needs some work.

[laxreference]

I wanted to put all the game preview links in one place for easy access. One day away. Happy Friday...

Saturday Games:
Loyola vs Syracuse
Yale vs Georgetown
Penn vs Army
Virginia vs Robert Morris

Sunday Games:
Penn State vs UMBC
Towson vs Maryland
Duke vs Richmond
Notre Dame vs Johns Hopkins

[Farfromgeneva]

Just curious if only the Hobart-Robert Morris final has been flipped would Hobart have moved up to 20 and would that have made enough difference for Cornell?

[laxreference]

And we are live with the win probabilities for the first game of the day, Loyola vs Syracuse. Enjoy

[CU77]

Just curious if only the Hobart-Robert Morris final has been flipped would Hobart have moved up to 20 and would that have made enough difference for Cornell?

No and no.

[laxreference]

Updated list of Live Win Probability Links

- Yale vs Georgetown
- Loyola vs Syracuse

[laxreference]

Updated list of Live Win Probability Links

- Penn vs Army
- Yale vs Georgetown
- Loyola vs Syracuse

[Hawkeye]

Updated list of Live Win Probability Links

- Yale vs Georgetown
- Loyola vs Syracuse

The parts of the Georgetown/Yale game where the curve is downward at a steady rate, are you missing data there?

[laxreference]

Updated list of Live Win Probability Links

- Virginia vs Robert Morris
- Penn vs Army
- Yale vs Georgetown
- Loyola vs Syracuse

[laxreference]

Updated list of Live Win Probability Links

- Yale vs Georgetown
- Loyola vs Syracuse

The parts of the Georgetown/Yale game where the curve is downward at a steady rate, are you missing data there?

No, not missing data. There are two parts to the estimate of win probability. Part of the calculation is the current time/score and what percentage of the time that the winning team holds on given those inputs.

The second part is the pre-game win probability, which steadily loses importance as the game goes on. It starts as 100% of the estimate and ends at zero of course.

During the flat line, the lead for Yale was large enough that the first part of the calculation essentially gave Georgetown a 0% chance to win. But since the pre-game estimates steadily degrade over time, that part of the estimate was moving slowly downward.

Put it another way, the volatility in the charts comes from the on-field plays as well as the time/score. Since there was no volatility on that piece, the chart itself ends up looking pretty boring.

[laxreference]

As always, here is the daily digest for May 11, 2019.

[laxreference]

The live win probability engine is up and running for Penn State vs UMBC.

[laxreference]

Updated list of Live Win Probability Links

- Towson vs Maryland
- Penn State vs UMBC

[laxreference]

Updated list of Live Win Probability Links

- Notre Dame vs Johns Hopkins
- Duke vs Richmond
- Towson vs Maryland
- Penn State vs UMBC

[laxreference]

As always, here is the daily digest for May 12, 2019.

[laxreference]

Here is the weekly list of the top 20 individual statistical performances of the opening round of the tournament.

Bucaro and Ierlan had quite the duel in their game. Granted, they didn't truly go against each other, but I think the fact that they took the top two spots says a lot about how that game went. The stats don't always isolate the factors that gave a specific game its character, but in this one, I think EGA did a pretty good job.

PlayerGame_visualization(20190513).png (54.49 KiB) Viewed 4135 times

[WildPride]

I think the face off rules need to be reevaluated when 5 of the top 8 performers are FOGOS.

[laxreference]

Updated list of Live Win Probability Links

- Virginia vs Maryland

[laxreference]

Updated list of Live Win Probability Links

- Duke vs Notre Dame
- Virginia vs Maryland

[laxreference]

Updated list of Live Win Probability Links

- Penn vs Yale