I have a general question about how the Sim handles pitching.  Not directly Dynamic Pricing related, but maybe could be an item to think about if this isn't already considered.

I've certainly noticed that low-strikeout pitchers tend to be a better bargain.  The opposing hitter makes contact earlier in the count than he would against a high-K pitcher, so the at-bat comes to resolution much sooner, making the low-K pitcher a much more efficient pitcher.  I'm not super-experienced on this game yet, but it seems the high-K pitcher fatigues faster.  I think this is partly why 1908 Addie Joss is so popular, he induces contact with low walk rates and low K rates, and gets through many at bats in only 3 or 4 pitches.  Meaning you squeeze a few more IP out of him and stretch your dollar.

But if 85 Doc Gooden made it through 270 innings with about 1K per inning for the Mets (or whatever it was that he did), while 89 Saberhagen made it through about the same number of innings but with only 0.6 Ks per inning, then wouldn't it stand to reason that Doc Gooden threw more pitches in 1985 than Saberhagen did in 1989?  What I'm getting at is that I don't think 85 Gooden should fatigue at the same rate as 89 Saberhagen.  To figure out the expected pitch count for pitchers on a season, does the Sim take into account differences in walk rates and K rates, or does it just figure 270IP is 270IP, no matter the K rate, contact rate, or walk rate?  I feel like 85 Gooden should be capable of throwing more pitches in a season than 89 Saberhagen, because reality is he probably did.  He worked deeper into counts to bring more at-bats to resolution.

WiS estimates pitches thrown, which takes into account how many batters the pitcher faced, as well as his K and BB rate. Search for the excellent post by elbirdo to see the details.

OK.  A step ahead of me for sure.  Thank you.

In practice, there is a small bonus for getting high-K pitchers. They are credited with more pitches to work with (as elbirdo noted), but they don't use all the extra pitches to get batters out. It's not a big effect, but it's there.

Posted by brianjw on 12/22/2015 2:33:00 PM (view original):
In practice, there is a small bonus for getting high-K pitchers. They are credited with more pitches to work with (as elbirdo noted), but they don't use all the extra pitches to get batters out. It's not a big effect, but it's there.

And I don't think it offsets how much they cost. 

Posted by bagchucker on 12/21/2015 2:32:00 AM (view original):
1908 Addie Joss is on 14% of all teams, 60% of all playoff teams, and 10% of all world series winners

there are something like 35,000 pitcher-years in the database

I suspect that these numbers, as well as the Bip Roberts numbers, are wrong.  In fact, I know they are wrong.

First of all, I doubt that if he's on 60% of playoff rosters, he'd only be on 10% of World Series rosters.  That would require the majority of all playoff teams without Addie Joss to win a World Series.  Seems unlikely.  But the big thing is that in OLs, 1 in 3 teams make the postseason.  If 14% of teams have Joss, at most 52% of playoff teams can have him on the roster.  If every single team drafting him makes it.

Ok, it's not a majority of playoff teams without Joss.  But it's a lot.  28%.  Whereas 2% of all playoff teams with Joss would be winning the Series.

Posted by dahsdebater on 12/23/2015 5:49:00 PM (view original):

Posted by bagchucker on 12/21/2015 2:32:00 AM (view original):
1908 Addie Joss is on 14% of all teams, 60% of all playoff teams, and 10% of all world series winners

there are something like 35,000 pitcher-years in the database

I suspect that these numbers, as well as the Bip Roberts numbers, are wrong.  In fact, I know they are wrong.

First of all, I doubt that if he's on 60% of playoff rosters, he'd only be on 10% of World Series rosters.  That would require the majority of all playoff teams without Addie Joss to win a World Series.  Seems unlikely.  But the big thing is that in OLs, 1 in 3 teams make the postseason.  If 14% of teams have Joss, at most 52% of playoff teams can have him on the roster.  If every single team drafting him makes it.

i stand corrected!

1908 Addie Joss on 14% of all open league teams, 25% of all playoff teams, and 36% of all world series winners
1992 Bip Roberts on 13% of all open league teams, 22% of all playoff teams, and 27% of all world series winners

i think my point was, there are 80,000 player-years in the base and the FACT THESE TWO GUYS dominate the postseason has to have a lot to do with mis-pricing

and of course also that they are yeomen at their cap

in a perfect world we would have all caps equally sampled by all users and all player salaries derived from the results

what we got here is schmeer

Joss on 36% of all Series winners is nuts. From 1961-1966 Sandy Koufax was on a smaller percentage of Series winners. 

But bagchucker, the market solution here may be just what we do need, since the more each is used the more they will cost and so that should reduce their use. 

Keep in mind that you are using a market - supply and demand - the total number of players is the supply and the demand is the huge percentage that uses those two players that stand in for cookies in general - so we actually did need a market to know that they were under-priced. 

Their pricing will now adjust itself. It can't happen in monopoly situations like the real world economy or MLB baseball where the players monopolize skills at a certain level and have a union, (a free market of teams would actually worsen the situation, expanding demand without supply growing), but it will work here. 

My concern is still that the market is based on the demand in a subset of leagues, but the supply will affect other leagues not included in that calculation.  It's like having the price of beef in Germany set by the market in the United States.  In 1700, when those markets didn't constantly communicate.