value calculators

[The Feral Slasher]

Maybe I'm not understanding the exercise but it looks like all you are doing is intentionally devaluing your pitching budget by increasing the number of pitchers. If that's what you want to do why not just reduce your pitcher %?

Isnt that what daily leagues do? They devalue "some" pitching while adding value to other pitching

[The Feral Slasher]

Maybe I'm not understanding the exercise but it looks like all you are doing is intentionally devaluing your pitching budget by increasing the number of pitchers. If that's what you want to do why not just reduce your pitcher %?

So reducing the pitcher percent would devalue all pitchers the same, it wouldn't add value to the pitchers who are not used in weekly leagues, but provide value in daily leagues.

[Ken]

So reducing the pitcher percent would devalue all pitchers the same, it wouldn't add value to the pitchers who are not used in weekly leagues, but provide value in daily leagues.

OK, but in your example, you are taking away $ from a player who you would never bench to give it to a player you only start part of the time. I don't understand why you would do that? Is that representative of their values? Is Degrom truly worth significantly less because you can start players like Patrick Sandoval several times? I don't see the correlation? But I may be missing it.

Also, you are valuing more pitchers than can be bought at auction, so in reality you are taking a $1 dollar days guy and saying he's worth $3 because of the league setup. But he's not, he's still worth $1, right? The last pitcher picked has to be worth $1, why would you pay more than a buck when there's no competition for that purchase?

[Ken]

Isnt that what daily leagues do? They devalue "some" pitching while adding value to other pitching

IMO the reality is that streaming has value in those 2x or daily leagues. Streaming involves using the marginal $1 players. It doesn't really mean those guys are "worth" more, it just means you should do it.

[The Feral Slasher]

OK, but in your example, you are taking away $ from a player who you would never bench to give it to a player you only start part of the time. I don't understand why you would do that? Is that representative of their values? Is Degrom truly worth significantly less because you can start players like Patrick Sandoval several times? I don't see the correlation? But I may be missing it.

Also, you are valuing more pitchers than can be bought at auction, so in reality you are taking a $1 dollar days guy and saying he's worth $3 because of the league setup. But he's not, he's still worth $1, right? The last pitcher picked has to be worth $1, why would you pay more than a buck when there's no competition for that purchase?

The top pitchers are worth less because they are smaller percentage of the overall statistics, and thus have less impact on them.

I'm not sure what exactly you are asking in the second paragraph. When there are daily transactions more pitchers are used and part of the statistics pool. That changes things.

[The Feral Slasher]

IMO the reality is that streaming has value in those 2x or daily leagues. Streaming involves using the marginal $1 players. It doesn't really mean those guys are "worth" more, it just means you should do it.

ok....

[Ken]

I'm not sure what exactly you are asking in the second paragraph. When there are daily transactions more pitchers are used and part of the statistics pool. That changes things.

Auction values are ultimately the % of your budget on auction day that you should allocate to a player.

Given that you only have 9 pitchers, that % can only be allocated to those 9 players (and 9 x number of teams total pitchers in the league).

If you are giving an arbitrary auction "value" to pitchers who don't get drafted among the top 9 x #ofTeams pitchers, then that value is not an auction value.

[chancellor]

The top pitchers are worth less because they are smaller percentage of the overall statistics, and thus have less impact on them.

Conceptually, definitely agree with you. And I'll readily admit I haven't dug hard enough into the math. But my gut read is a 35% value drop for the top-tier pitchers is too much.

Interesting look at how to manipulate the calculator, though. Would love to see the RPs if you run them.

[Ken]

The top pitchers are worth less because they are smaller percentage of the overall statistics, and thus have less impact on them.

I get the concept on counting stats, on the ratios, their ratios are going to be more valuable because they are further ahead of "replacement" when marginal pitchers are given more innings (assuming you are pitching more starters and not MR).

[The Feral Slasher]

Conceptually, definitely agree with you. And I'll readily admit I haven't dug hard enough into the math. But my gut read is a 35% value drop for the top-tier pitchers is too much.

Interesting look at how to manipulate the calculator, though. Would love to see the RPs if you run them.

yeah, 35 % may not be the right number...like I said I took about 15 minutes to just think up some assumptions, clearly changes could be made to refine this. The intent was to have the calculator reflect how the leagues actually function and how many pitchers of each type are actually used over a week, not just how many lineup slots are assigned. I'll run the relievers at some point and post. I expect MR should see a big bump in value- they offer little value in a weekly league in general, but in the one daily league I have i start several MRs 4-6 times a week. So I hopefully the results will match the reality

[The Feral Slasher]

I get the concept on counting stats, on the ratios, their ratios are going to be more valuable because they are further ahead of "replacement" when marginal pitchers are given more innings (assuming you are pitching more starters and not MR).

In my daily league i use MR guys all the time. Why would people not do that ?

[The Feral Slasher]

Conceptually, definitely agree with you. And I'll readily admit I haven't dug hard enough into the math. But my gut read is a 35% value drop for the top-tier pitchers is too much.

Interesting look at how to manipulate the calculator, though. Would love to see the RPs if you run them.

I'm just going to do the weekly vs daily comparison. First column is for daily league, 2nd column for weekly. Perhaps not as drastic as the starters.

Edwin Diaz 17.2 21.9
Hader 14.9 18.3
Chapman 14.8 17.9
Hendrik 14.8 17.9
Kimbrel 13.1 15
10- 10.2 9.8
31 - 6.2 3.9
61 - 3.2 -1
91 - 1.9 -3.1
121 - 1.0
151 - 0.2

[The Feral Slasher]

I'm just going to do the weekly vs daily comparison. First column is for daily league, 2nd column for weekly. Perhaps not as drastic as the starters.

Edwin Diaz 17.2 21.9
Hader 14.9 18.3
Chapman 14.8 17.9
Hendrik 14.8 17.9
Kimbrel 13.1 15
10- 10.2 9.8
31 - 6.2 3.9
61 - 3.2 -1
91 - 1.9 -3.1
121 - 1.0
151 - 0.2

so the formatting is crappy, but one conclusion would be that about 120 relievers should be rostered in a daily league, which actually agrees (or is closer to reality) with how i suspect those leagues usually work (mine does). Using the values from the same inputs as a weekly league would indicate that the 60th reliever off the board is providing negative value. So why would anyone bid on or start a guy with negative value ? Yes it could happen with streaming, but that isn't typical of middle reliever usage that i have seen. they are used as guys to fill in and get a small amount of counting stats when there are not decent starters available. Anyway, I'm sure what I have put together isn't perfect, but it appears to me to be a very significant improvement over using the same inputs in daily and weekly leagues.

[The Feral Slasher]

Auction values are ultimately the % of your budget on auction day that you should allocate to a player.

Given that you only have 9 pitchers, that % can only be allocated to those 9 players (and 9 x number of teams total pitchers in the league).

If you are giving an arbitrary auction "value" to pitchers who don't get drafted among the top 9 x #ofTeams pitchers, then that value is not an auction value.

I think you are not correct here. You are not accounting for reserves.

What I am giving value to is pitchers who produce stats that generate value. In order to determine how to allocate value the first thing that needs to be done is to figure out what the total amount of wins, K's, saves will be and what the ERA and WHIP are. Otherwise it is not possible to calculate the value of the stats each player generates. This can't be done accurately for leagues with daily transactions if you assume only 9 pitchers are generating stats, because that isn't how it works in practice. Using the same calculator for daily and weekly leagues will fail because it won't accurately calculate the total stats that are generated and it won't allocate them appropriately. I wouldn't think this would be controversial.

Anyway, there are still some adjustments that need to be made for streaming pitchers and relievers that are likely to be in the lineup every day vs 1-5 days a week as well as league preferences historical bids, etc. But the method I proposed gives you a much better starting point to make adjustments from because it more accurately predicts the total stats for the team/league and distributes them among a wider pool - which is what happens when you have daily lineup changes instead of weekly.

[Ken]

I think you are not correct here. You are not accounting for reserves.

What I am giving value to is pitchers who produce stats that generate value. In order to determine how to allocate value the first thing that needs to be done is to figure out what the total amount of wins, K's, saves will be and what the ERA and WHIP are. Otherwise it is not possible to calculate the value of the stats each player generates. This can't be done accurately for leagues with daily transactions if you assume only 9 pitchers are generating stats, because that isn't how it works in practice. Using the same calculator for daily and weekly leagues will fail because it won't accurately calculate the total stats that are generated and it won't allocate them appropriately. I wouldn't think this would be controversial.

Anyway, there are still some adjustments that need to be made for streaming pitchers and relievers that are likely to be in the lineup every day vs 1-5 days a week as well as league preferences historical bids, etc. But the method I proposed gives you a much better starting point to make adjustments from because it more accurately predicts the total stats for the team/league and distributes them among a wider pool - which is what happens when you have daily lineup changes instead of weekly.

I'm not sure if its possible for this conversation to be useful but I'll try anyway, I'm a glutton for punishment.

You are correct I'm not accounting for reserves. By definition. That's how auction $ calculators work. They are telling you how much $ you should spend so that you do not have to be stuck with a replacment level player (i.e. one you can get in the reserves or FAAB).

The concept is predicated on some assumptions, that while not perfect, are close. Conceptually the "talent" curve is asymptotic. It's steep on the high $ end and flat on the low $ end. For the purposes of illustration we can call the BEST player that isn't picked in the auction our replacement level player, and assume that every replacement level player is the same. While this is a slightly flawed argument, we can reconcile it with an understanding that if every team grabs replacement level players and those players on average perform around the same, the actual stats of the replacement level players don't matter since they net out. For valuation purposes they are ubiquitous.

Let me provide an example. Lets say we start with a weekly league that starts 9 pitchers. And for the purposes of illustration lets say we have established that Spencer Turnbull is the best pitcher who would not get drafted.

When we switch to daily or 2x transactions, lets say that every team basically adds a bunch of "Spencer Turnbull" starts, over and over. Heck lets take it to the extreme and go ahead and GIVE 10 Spencer Turnbulls to every team. So we can turn this back into a weekly lineup with our normal 9 starters and then 10 Spencer Turnbulls on every team.

The standings are exactly the same at the end of the year with the 19 pitchers (10 of which are Spencer Turnbull) as they are with our normal 9. The Spencer Turnbull stats all net out.

So Jacob Degrom's value isn't relative to the total stats that are accumulated, his value is relative to the total stats MINUS all those replacement level starts that every team is getting.

If you are placing positive $ values on players that are "reserves" in your own valuations, then IMO you are incorrectly anchoring your replacement value.