auction drafting in mixed leagues -- closer pricing?

[mjlewis]

I've been comparing two auctiony-pricey-comparey-tools (MB and Baseball Prospectus) and one of the big differences between the two comes in pricing of closers.

Using a 10-team mixed league, MB has the best closers coming in around $12 (where the best starters are $30 or so) and the 10th-best around $6. Baseball Prospectus has the best closers at around $18 (with the best starters still at about the same prices) and the 10th-best around $10. Do you guys have an opinion of which of these pricings is more "correct"? I can see the MB component pricing and not the BP, unfortunately, but I can see that both of their projections for player performance are... not exactly the same but not far enough apart to explain those pricing differences.

- Mike

[viper]

I'm sure Todd can detail the reasons from a mathematical point of view. From a practical point of view, every league, meaning group of people, value closers differently. This is a case of know thy league.

I'm in a league where closers are almost hated., There are several who would like to eliminate the saves category. MRs go for maybe $2. I took Thorton for $3. Not because i had to bid that high but I had $3 left and he was my last player. One other owner could go to $3 so I nominated him for all my remaining money. I really wanted him as I had Jenks as a closer.

When i spread dollars before the draft, I allocate $20 for two closers and one MR. The top guys go for $15-$20 and I look for two lesser closers in this AL-only league.

[rotodog]

Want my take? the correct price is what your league pays for them...Simple.

If your league pays 5-13 bucks for closers, than thats the right price....If your league pays 10-20 for closers, then that is the right price.

Use the dollar figures as basically a ranking tool... If closers are going in the 10-20 range, you can then adjust the top tiers towards 20 and the lower tiers toward 10 and adjust in the middle.

But neither one is "correct". The correct price depends on what your league pays for closers...So dont use one or the other as a ceiling to bid..Find the value in context of the rest of the closer prices in the league..

FWIW, I have had access to BP pricing engine in the past....An old partner was a BP and Baseball Notebook freak....We had many disagreements on pricing in the past. I won most of them fortunately.. Both pricing tools are lacking..I never like BP projections and I hated the BB notebook pricing calculations and 80% rule that his readers always employ incorrectly..

if I had to choose one, it would be MB calculations. At least I know the basic method in which they are constructed..I can tweak a price when i know how it was calculated much better than other tools..

[mjlewis]

of course there's value in knowing how the rest of the league thinks, but I am pretty sure there is a canonical right answer here. It may be too difficult to figure out that right answer, especially as the team balance changes during the auction, but it exists. If everyone else is paying $30 for closers, that doesn't mean that I should -- it probably means I should punt saves. If everyone else is paying $5 for closers, maybe I should take an extra one.

I appreciate the concept that the right answer is somewhat league-dependent in terms of what you do with the numbers, but there is a right number, and if we've learned anything from Moneyball, it's that everyone else might be wrong and you can find the relatively underpriced option among the other choices.

And if you say something like "if closers are going in the 10-20 range, maybe you wait to strike on the ones that will cost 8-10 dollars", isn't that saying that closers at 10-20 are overpriced? The price they'll go for is what your league pays for them, sure, but their value isn't.

edit: an addendum: and note that the pricing difference I was referring to is not relative between closers -- it's that one pricing model has all closers, and saves in general, more expensive than the other one.

- Mike

[viper]

OK then, let's assume you plan a 180/80 mix with $80 for pitchers. I guess you would come up with a starter/non-starter mix that you like and proceed accordingly. In my league I planned to split the $80 on a $60/$20 basis. I guess you have to do something similar.

An interesting exercise would be to see the total dollars your other site has for all positive valued pitchers. It sound like it will exceed $800 which would be $80 time 10 teams.

[rotodog]

of course there's value in knowing how the rest of the league thinks, but I am pretty sure there is a canonical right answer here. It may be too difficult to figure out that right answer, especially as the team balance changes during the auction, but it exists. If everyone else is paying $30 for closers, that doesn't mean that I should -- it probably means I should punt saves. If everyone else is paying $5 for closers, maybe I should take an extra one.

I appreciate the concept that the right answer is somewhat league-dependent in terms of what you do with the numbers, but there is a right number, and if we've learned anything from Moneyball, it's that everyone else might be wrong and you can find the relatively underpriced option among the other choices.

And if you say something like "if closers are going in the 10-20 range, maybe you wait to strike on the ones that will cost 8-10 dollars", isn't that saying that closers at 10-20 are overpriced? The price they'll go for is what your league pays for them, sure, but their value isn't.

edit: an addendum: and note that the pricing difference I was referring to is not relative between closers -- it's that one pricing model has all closers, and saves in general, more expensive than the other one.

- Mike

Actually, with a name like MJlewis I should have seen the Moneyball reference coming...great book by the way.. I am dying to read his new book "The Big Short"..But for some dam reason, it is not available for my Amazon kindle...I was p*ss*d when i got home from Barnes and Noble Sunday night and I went to buy it on my kindle..

Back to baseball.....

I wasnt trying to give wiseguy advice. Or even anecdotal advice....

The concept I was trying convey is that some systems will overvalue saves and some may not...I dont think in context one is right and one is wrong. If your league pays within a certain level for closers, your job is to find the value within that context if you want saves..if they are paying 30 bucks for closers, then yes, maybe you punt..

But from a mathematical correctness point, I trust the MB way to valuate stats properly. The method they use is probably more mathematically correct than some other systems as it is based on replacement level and values everything in context from replacement level on up.

I do not know how BP is now creating there values. BP was always a great Sabermetric commentary, but in the past I have found there projections and valuations for the purposes of fantasy to be substandard ...They are baseball Statheads for the most part and not fantasy baseball experts.. I know they have added to there offerings in the last couple years to cater to the fantasy market, but I still think there projection metrics and valuation process is not as good as others..Mastersball included...

Todd is the valuation guru here...I am sure he can find a Mathematically correct answer for you, but it still shouldnt override the abilty to buy players and stats that you need to win a league.

If you do get a satisfactory answer that is mathematically correct and it suggests that MB valuations are more correct than others, but the rest of the league doesnt get that memo, you may not buy any saves because you will think they are all overpriced... If you were satisfied with the MB values and used them as a bid ceiling and stopped bidding for a closer at 9 bucks that everyone else is paying 12 bucks for, you might get shut out..

So thats why I mention using the valuations to identify the value within the context of how the league pays for saves...

As Billy beane and Paul Depodesta did in the early part of this century, they identified value within the context of the whole league..That was OBP and certain mathematically correct ways to play the game....Once everyone else started doing it, there was no value within that context....So they went to Run prevention...Now Theo and the young bunch is now paying for defense and run prevention...Once Defense begins to become overpriced within context of the league, they will find some new way to exploit the system...

What matters more than Absolutely correct values is roster construction and budgeting....And then paying for the stats you need to win...this includes spending a bit more to get it sometimes and sometime it means identifying underpriced assets at others..But it is all within the context of what others are paying on draft day..

Did any of that make sense? I wrote it, and I am not sure it does...

[mjlewis]

It made sense, and I agree. (I've been kind of dubious of BP's fantasy tools for the last couple years but this is the first time I actually gave up on them and moved to a different platform.) I know that one of my biggest weaknesses in the auction is in figuring out when to let go of my pricing preconceptions and get the guys I need to make my team work.

And my name is actually Michael Lewis but it's just a coincidence. Once I was on a date and I had one of his books in the back seat and my date thought I was trying to show off about having written a book. I had to get the book and show her the jacket picture to convince her it wasn't me

- Mike

[rotodog]

It made sense, and I agree. (I've been kind of dubious of BP's fantasy tools for the last couple years but this is the first time I actually gave up on them and moved to a different platform.) I know that one of my biggest weaknesses in the auction is in figuring out when to let go of my pricing preconceptions and get the guys I need to make my team work.

And my name is actually Michael Lewis but it's just a coincidence. Once I was on a date and I had one of his books in the back seat and my date thought I was trying to show off about having written a book. I had to get the book and show her the jacket picture to convince her it wasn't me

- Mike

I would suggest that you( if you havent already) find the main Mastersball homepage and go read the 2-3 pieces Todd has written for SI.com . They deal specifically with things I touched on here... I used a program called Rotolab for a few years and gave it up 3 years ago for some of the same reasons I mentioned. I now love showing up to a draft with guys using rotolab or other computers as I know they tend to stick tight to what the computer is spitting out in front of there faces... If you identify the guys you need to win and put together a solid plan, a couple bucks here or there isnt going to alter it much.... Just go get the guys you want and fill in....

But, I am sure Todd will expand further on the mathematics tutorial at some point....But as a scientist and mathematical valuation guru, he tends to look for mathematically sound conclusions with no room for tolerances...That said, I know he has abandoned the absolute correct valuations stand and may push you towards other ways to win a league..

Good Luck..

[JP Kastner]

Trying to figure out when to ignore the values and go with your gut is one of the hardest things to explain. In the Tout Wars Mixed League, there is a premium on catchers and closers. I know that I'll have to pay more for them, so I budget more for those slots. Some times budgeting per roster spot is enough to compensate for known league tendencies.

The technique I use to assist me is the one Lawr taught me years ago. I figure out how many statistics I need to win and I make a list of players that will achieve that. Once I have a good idea of who those players are, I draw lines on my list of players broken down by position. As long as my team is made up of players over that line, I am in good shape.

Again, using this year's Mixed Tout, I knew my strategy required that I pick up some power later at the cost of AVE. I budgeted $30 for my #1 outfielder. When Ichiro Suzuki's name was called, I knew that he would give me two things I needed badly at that point, SB and AVE. Our projections have him at $22. My slot was for $30. I paid extra for him ($26). I was then able to spend a $1 for Jack Cust's 24 HR and not kill my average. My goal was not Cust, but Jay Bruce, but Bruce went to way more than what it would take to make some profit.

The combination Suzuki and Cust is .291 with 16 HR and 15 stolen bases for $2 profit. I gave a little but gained in the end. Our projections on Suzuki are conservative, so he might make $26, then my profit is even higher.

[daweasle]

another important factor in this is the timing of the draft

for example early on in my recent mixed league draft the first 7-8 top closers to go off the board went for $17-22 (krod wagner papelbon rivera)

then a lull when guys like mike gonzales and bobby jenks went for $3-4 each

then when closers started getting scarce - they went for $12-15 again.
(rafael soriano and brian fuentes)

and near the end there were 2-3 closers (with the title) who went for $1-2 (frasor/kerry wood/lidge/someone else)

timing is a huge factor in auctions, and if you can somehow help control timing - you can also help control prices.

(Let's say for example when it's time for you to nominate your fist player - you throw out kerry wood for $1 - that's zigging when everyone else is zagging) everyone will be shocked and you just might get an established closer for $1 - or someone else might buy him for $10 - but i bet you it will get a reaction and someone else might just throw out another random lower tier closer before it's your pick again.

[alleyoops]

Hi Mike - good to see you're thinking about this topic - especially since I'm trying to peddle some closers in that 10-team league.

I agree with you - there should be a right or wrong answer to the valuation question. I don't use BP, and have never seen their projected values. I think MB's are somewhat in line with what has been the case in the shallow mixed leagues in which I've played (including this one). Historically, in the early years, the league overpaid for closers, but in recent years it has been the opposite. The trend recently is for the best closers to go between 15 and 20, and the rest to go for single digits. Dropping from 12 teams to 10, I suspect they'll all drop some in cost, or at least not cost more than they did last year. That's cost though - not value.