Announcement

**WillyBoy** · 05-04-2019, 03:17 PM

The Limper - NBA 2019 - Playoffs - 5-4.png

**WillyBoy** · 05-05-2019, 02:10 PM

The Limper - NBA 2019 - Playoffs - 5-5.png

**WillyBoy** · 05-06-2019, 12:39 AM

That’s 5 straight-up losses in a row!
It’s one thing to lose against-the-spread, but picking game winners is something the model should, at the least, be able to do reliably. That’s especially true of the playoffs, when all the data and trends are available, and players are supposed to be playing at their peak performance levels. The numbers should be a reliable basis for accurate projections, but it’s not working out that way. Prior to the last few days, the model did have a good run, and probabilistically speaking, a regression was inevitable – but it flat out fell off a cliff, which is downright bizarre.
Anyway, proceed with caution.

**BettorsChat** · 05-06-2019, 08:59 AM

Originally posted by WillyBoy View Post

That’s 5 straight-up losses in a row!
It’s one thing to lose against-the-spread, but picking game winners is something the model should, at the least, be able to do reliably. That’s especially true of the playoffs, when all the data and trends are available, and players are supposed to be playing at their peak performance levels. The numbers should be a reliable basis for accurate projections, but it’s not working out that way. Prior to the last few days, the model did have a good run, and probabilistically speaking, a regression was inevitable – but it flat out fell off a cliff, which is downright bizarre.
Anyway, proceed with caution.

Do you ever tweak the system that you have even during the playoffs?

**WillyBoy** · 05-06-2019, 02:30 PM

Originally posted by BettorsChat View Post

Do you ever tweak the system that you have even during the playoffs?

Except during the playoffs, I’m tweaking the damn model all year long - LOL. The only changes I make for the playoffs are the normal player substitutions. I do recalculate player trend and past performance data, based on previous playoff experience, but with only a handful of exceptions, there’s not much that changes.

Based only on the eyeball test, however, I think I’ll have to recalculate officiating performance numbers. It’s making more and more of a difference, as I think officials are more impacted by the playoffs stage they’re on than I had anticipated.