I've been leaning on it more and more. Can't say it's helped. Thoughts?

I'm pretty new and I probably have looked at it more than I should, but I think if there have been a lot of seasons played then it is probably a reliable sample size. I notice some have very few seasons, and worse than that is when there is a really aberrant best or worst season that skews the numbers significantly.

I suppose it would be good to know if it just uses open leagues for those, or if crazy themes are skewing the stats.

Seems like only open leagues get recorded in performance history. Maybe someone who knows for sure can answer?

Anyway, relying on performance history is like relying on past trends on the stock exchange. A decent indicator, but not too reliable. Stick with the advanced stats instead.

Performance history only counts open leagues.

Posted by darthmetool on 1/8/2017 11:02:00 PM (view original):
Seems like only open leagues get recorded in performance history. Maybe someone who knows for sure can answer?

Anyway, relying on performance history is like relying on past trends on the stock exchange. A decent indicator, but not too reliable. Stick with the advanced stats instead.

You may want to consider a course in statistics.

https://www.whatifsports.com/forums/MakePost.aspx?TopicID=422705&QuotePostID=9074301

Part of a previous discussion about performance history.

it's useful for some stats, not useful for others. I look at A+ and D- range fielders to see how many + or - plays they've made because these ratings have a wide spread. Also an A+ arm Catchers CS% is a useful indicator. looking at HR numbers across the years gives me an idea how power will normalize. But for most stats PH is not very reliable because you don't know how they were used or abused. RP numbers are particularly bad because of abandoned teams leaving them in to get torched. So bad Performance History doesn't mean much but a player used often who performs better than you expected can be meaningful.

Ditto as far as a catcher's CS percentage and a player's fielding percentage. However a player's PH can be helpful in other was as well. For example, the guy may have a 3# normalization rating for triples, but despite being used quite frequently, he only averages 10-12 per year. This is a VERY good indicator that despite his high triples rating, he will not hit very many.

Is there any useful explanation as to why a player's normalized stats would not result in, as in the above example, the expected number of triples?

Posted by redcped on 1/13/2017 2:09:00 AM (view original):
Is there any useful explanation as to why a player's normalized stats would not result in, as in the above example, the expected number of triples?

The "#" numbers are just predictions. The "+" numbers are actual stats. So if someone has triples at 3/100#, that's just a predicted normalization by the computer, but it's not like 100% correct cuz it's not an actual like, real stat

But the normalized stats are what the game uses, not the real ones, correct?

Performance history can be useful as a factor in who to draft, but shouldn't be the ultimate authority. A lot of times, when I've narrowed a pick down to a few people, that's when I'll look at performance history as a virtual tiebreaker when players seem very similar.

Great explanation d_rock97!!!!