Draft Pick Compensation Revisited

[cabot41]

After becoming frustrated with the way OOTP implements draft pick compensation for lost free agents (a frustration that I have expouned on elsewhere) I actually hunkered down with the odious MLB rulebook to see how the big boys do it.

Not surprisingly it is enormously complicated and I don't think that any "person who likes playing games" really wants to deal with the "realism" of hiring a team of lawyers to figure it all out each free agent season. Still and all, I think the current "Cliff notes' version that ootp uses is deeply flawed, in essence punishing bad teams that try to improve through free agency, which is the manifest oppoosite of what the system is trying to do.

I think that draft pick compensation can be implemented in this sim, in a basically accurate way, without making gamers want to stab out their eyes.

My suggestions: A) IRL the top 15 (top half of 1st round) picks are protected and this simply HAS to be implemented in the sim. Anything less simply punishes bad teams for trying to go out and sign the best FAs to help their club. If this doesn't happen, everything else is moot.

B) The heart of the MLB system is actually not swapping picks b/w teams, but the sandwhich rounds. At base the sandwhich rounds are not complicated concepts, its just a round that comes after the first, before the 2nd, and so on. Teams get sandwhich round picks only IF they loose compensation eligible free agents, that is why the sandwhich round has an "undetermined" number of picks. In one year there might be 10, in another 20. These rounds are necessary because the sandwhich rounds solve the problem of 1 team signing FAs from 2 different teams. Every team that looses a compensation worthy free agent is supposed to get ... compensation. Without sandwhich rounds, this can't happen because the signing team might have already given away their draft choice. So to make this work, we need to create "extra picks" which is exactly what the sandwhich round does. Two drafts ago the Red Sox got a slew of sanwhich round picks for loosing Pedro, Lowe, Embree, even Doug Man-Cant-Hit. Pedro and Doung both went ot the Mets, and Boston still got two picks, even though the Mets also signed Beltran AND had their first round pick lottery protected (which they used for Mike Pelfry). We need sandwhich rounds.

C) Where MLB gets out of controll complicated is determining who gets another team's pick when it is not protected and that team signs multple FAs, and determining the order of the sandwhich round selections. But we can scarifce a little reality to obviate that complication in two simple steps

(Continues in next post)

[cabot41]

(Continued from Above)

C-1) Limit the number of Free agents that are eligible for draft pick compensation.
IRL, MLB has a bunch of factors taken into account to determine if a FA is a type A, B or whatever. If a player meets whatever statiscal threshold, they are designated as compensation eligible and that is the end of it. In the sim, why can't we say put numerical cutoffs on on the amound of FAs that are designated for compensation? Off the top of my head, it seems like we could just limit the amount of tyep A FAs available to say, 5. The five best FA players are eligible for 1st or 1/2 sandwhich round compensation. THe sixith best free agent ... well he gets bumped down to type B. Does that mean that some years the Carl Pavanos of the world are type A and other years they are type C, ... yes it does. But I think that sacrifice in realism would allow us to perserve the heart of why teams get compesation for FAs at all, yet massively reduce the complication of having multi-tiered tie breakers to determine sandwhich round draft order. If you only had 5 Type A FAs, even if one team signed all 5 of those guys, you would have a maximum of 5 first/second sandwhich round picks (assuming the signing team had their first rounder protected).

Having only 5 additional picks here and there also helps the programming nightmare of having to generate additional players to fill out a draft.

C-2) To limit the number of free agents that are eligble for compensation, you have to teach the AI to rank them. You have to teach the AI to say that BJ Ryan is better/worse than Billy Wagner (obviously, reasonable people will disagree about the distinctions). Once you have a rankings established, that becomes your basis for the tie breakers. The team that looses the Number 1 free agent player gets the swapped draft pick or the first pick in the first/second sandwhich round. The team that looses the Number 2 free agent player gets the swapped pick OR (if that pick has already been swapped for the No. 1 guy) the first pick of the sandwhich round OR (if that pick goes to in exchange for the No. 1 guy) the second pick of the sandwhich round ... and so forth.

If gamers can be taught the complications of the rule V draft, option years, and waiver rules, I think that gamers could follow this dumbed down version of the actual MLB draft rules. And then this feature, which is a great idea, could be made to do what it is supposed to do.

So ... what do you think?

[Syd Thrift]

Offering a player arbitration before he tests the market is an absolute necessity for making free agency work right. IRL, for a team to get a compensation pick they have to offer him arbitration. This is a big deal because sometimes a bad player will accept this rather than test the market and boom, you have to pay a guy you didn't want. The whole point of compensation ought to be to give a team something back for losing a player they didn't want to lose.

It's also part of baseball, too, that you can't trade off your picks, but since virtually every other American sports league allows this, what the heck... including it as an option is a'ight. I have no idea whether the computer values picks too much or too little because I have no real-life comparison to go by, but hey.

Anyway, I hear that next year we'll be able to find *all* players by sending scouts out to scour the country. That's going to be *huge* for pre-draft leagues.

[cabot41]

I forgot about arbitration.

Yes, this too is very important. You shouldn't get compensation for players you dont want.

[cabot41]

Do we know if the team is looking at making any changes to the draft compensation system at all? Its not nearly as important as some other things since it can be turned off, but the flaws kind of render the feature pretty unusable in fictional solo leagues.

[cabot41]

Polite bump.

[tysok]

One of the problems you have to work out as well is making it so you can't trade a free agent signed player right away. Nor an arbitrated contract.
One problem I've ran into, I had a bunch of money available and needed a good starting pitcher. There wasn't any good starters in free agency, so I signed 2 of the best OFs available, and traded both for a top notch starter. In reality you can't trade a newly signed free agent, I believe I read somewhere that every free agent contract comes with an automatic no trade for the first year... I should not have been able to do this. Not sure how arbitration contracts are handled for this purpose.

The reason you'll let this free agent walk, well one of them, is you can't afford him and know it. Arbitration will end up with a big contract. If you offer, and he surprises you by going to arbitration you can then just trade him off... probably for better quality than a hit-or-miss draft pick. You already know what he's going to make, (the same with the Astros and Clemens last year) pretty much whatever he's asking for in extension in the cases of these star players.
One case of a player accepting the arbitration instead of free agency is Maddux with the Braves, it pretty much forced Atlanta to trade Millwood before he could go to arb an get another huge contract they couldn't afford. Don't remember the site, but found this mentioned somewhere when I was doing a search on arbitration.

Of course this could be brought about by a honor rating, or whatever it might be called. Where you lose some prestige or honor by screwing these players over like this. Where future free agents, and even your own players in cases of extensions, jack up their prices or refuse to deal with you because you're a dirt bag liar etc...

[cabot41]

I obviously agree with the general sentiment, but I'm pretty sure that baseball is unique in major american sports in that it does not have a "defacto no trade clause" for recently signed free agents.

I cannot think of a good example off the top of my head, and I am sure it happens exceedingly rarely for big budget signings, but if you think about all the lesser moves teams make in the offseason, (signing pitcher X to the major league minimum not sure if he'll even make the club, then he does and is a usefull chip at the trade deadline) I'm sure we could find such examples. Clemens signed with the Stros after not being offered arbitration, (so it was a FA signing and not a "re"signing) and already there is talk about him being moved to the Yanks or Sox.

I think the reason real clubs can't exploit the system as you mentioned (signing the outfielders to trade for the pitcher) is because IRL agents can look at a depth chart as well as anybody else. No agent is going to let his stud OF sign in a place where he is not garunteed a starting spot in hopes of getting traded potentially to someplace his guy doesn't want to go.

I think the big point that I missed initially is that offering arbitration to elligble free agents is also crucial to making draft pick compensation work correctly in this sim.

[tysok]

"Herein lies the risk for the team in offering arbitration. Let’s use one of the premier free agents on the market this year, Marlins pitcher A.J. Burnett, as an example. By the time December 7 rolls around, Burnett will have had several weeks to scour the free agent market and get a sense of how much money he could get from other teams. His agent also has a pretty clear sense of how much Burnett could make if he went to arbitration. The Marlins have already made it clear they do not intend to re-sign Burnett. But they could still offer him arbitration to get those coveted draft picks as compensation from the team with whom Burnett signs. But what if Burnett accepts arbitration? What if, after playing the market, he’s not seeing the kind of money thrown at him he was expecting and he thinks he can get a fat contract in arbitration and test the market again next year? In that case the Marlins would be stuck with a player they don’t want and can’t afford. Since Burnett will almost certainly sign with another team for big money, the risk is low for the Marlins to offer arbitration. But the risk is still there. "

http://groups.msn.com/baseballforum/...tionrules.msnw
"Major League free agents come with an automatic no-trade clause until after the next June 15."

There it is. It's not the whole year, but enough to make you deal with it for a while. And would negate my signing of those OF, since I'd still be short a starter for 2 1/2 months, but carrying big contracts like I had one.

I don't know what rules come about with arbitration though, maybe very little since it's just a rent a player deal in that situation. But if it were in game I wouldn't care about the "risk" they mention, for reasons I mentioned. If I get a draft pick great, if I end up with him I trade him off for proven big league talent.

[bulldog55]

Some previous discussion on the matter (including my own thoughts):

Best thing I've seen so far - Compenstation Picks

[Elendil]

While modelling MLB fully is likely to be very difficult, it might not be difficult to implement just one feature: if a team with a top-50% (top 15 in a league with 30, e.g.) pick signs a free agent, it cannot lose its first-round pick; instead, it loses at most a second-round pick.

[cabot41]

I think bulldog's suggestion (as stated in his linked post) that you lett he users decide how they want to compensate types A and B and so forth would be excellent. Customization is always cool.

In the meantime, I've experimented with leaving compensation on but resetting the top half of the first round to the teams with the worst record, kind a manuel work around to make sure those top picks are protected. In my experience, it still just doesn't work right without the sandwhich rounds. Without sandwhich rounds teams that have their free agents signed later in the period get nothing. Swapping only works so long as the signing team has picks, so if you've got multiple compensation worthy free agents, you are left just "hoping" that they are all signed early and by different teams. It is just not fair.

So I'll say again, without adding sandwhich rounds, this feature will just not work correctly.

One theoretical question, how would you teach the AI to regard the cost/benefit of the compensation. IRL teams are all over the map: High Revenue teams sign who they want w/o much regard to the lost picks, Moneyball teams like OAK and BOS try to actively stockpile draft picks Bill Bellichek style. Then there are GMs like Brain Sabean who actively tries to avoid getting high draft picks (because he thinks that the investment in terms of signing bonus often outstrips the worth the risky proposition that is inherent to the draft). So I wonder, if implemented correctly, if the AI would adopt one valuation method, or if they would try to code the AI to have different responses depending on team revenue, win expectation, and so forth.

[Curtis]

Originally Posted by cabot41

(Continued from Above)

C-1) Limit the number of Free agents that are eligible for draft pick compensation.
(snip) If you only had 5 Type A FAs, even if one team signed all 5 of those guys, you would have a maximum of 5 first/second sandwhich round picks (assuming the signing team had their first rounder protected).

Having only 5 additional picks here and there also helps the programming nightmare of having to generate additional players to fill out a draft.

C-2) To limit the number of free agents that are eligble for compensation, you have to teach the AI to rank them. (snip)

If gamers can be taught the complications of the rule V draft, option years, and waiver rules, I think that gamers could follow this dumbed down version of the actual MLB draft rules. And then this feature, which is a great idea, could be made to do what it is supposed to do.

So ... what do you think?

I think your concept is great, but a simple way to get around 'the programming nightmare of having to generate additional players to fill out a draft' is to simply increase the size of the draft pool. I flatly couldn't believe it when I did my first amateur draft and the number of picks available (1000) exactly equaled the number of players available to be picked. Wouldn't it be great world if every player who wanted to be drafted actually was? (Sarcasm intentional.)

The logical way to run an amateur draft would be to have the pool be 20 to 25% larger than the number of picks available, and anyone who doesn't get picked becomes an instant free agent. That way, unless there was an astronomical number of sandwich picks, no additional players would have to be created.

This has a correllary with other positive aspects, namely that player retirement algorithms could be adjusted to favor retiring guys who haven't played in the past year. I was so pissed off when the four best pitchers on my rookie league team all retired at the end of last season. Two of them were 37 and 38, so I couldn't really complain, but another guy was 34 and had pitched 25 shutout innings, and the fourth was only 28 and had pitched 13.2 PERFECT innings! I guess he decided he'd climbed the mountain and it was time to go home, but I was planning to promote all four of those guys to single A this year, and now I have to start all over. I'd much rather have the computer save AT LEAST the 28 year old and retire some guy in the free agent pool who couldn't catch on with anyone.

Back on topic, I support cabot41's compensation pick suggestions.

[Bay Area Bob]

Uh, I think it's extremely realistic for anyone above 25 to retire if they're still in rookie ball. I have tremendous issue with all of these 30 year olds in rookie ball and single-A teams that are affiliated with big league teams in OOTP6.

[Le Grande Orange]

Originally Posted by Bay Area Bob

Uh, I think it's extremely realistic for anyone above 25 to retire if they're still in rookie ball. I have tremendous issue with all of these 30 year olds in rookie ball and single-A teams that are affiliated with big league teams in OOTP6.

That's true for current-day minor leagues. In earlier times, it was different, and players could make a living playing only in minor league ball. Also, there were a good number of major league players who, when no longer wanted at the major league level, went down to the minors and played several more years there before finally retiring for good.

I don't want to see the ability to mimic the earlier eras of the minor leagues sacrificed to solve issues with current-day minor leagues. I'd much rather have the ability to recreate both kinds.

[tysok]

Originally Posted by Le Grande Orange

That's true for current-day minor leagues. In earlier times, it was different, and players could make a living playing only in minor league ball. Also, there were a good number of major league players who, when no longer wanted at the major league level, went down to the minors and played several more years there before finally retiring for good.

I don't want to see the ability to mimic the earlier eras of the minor leagues sacrificed to solve issues with current-day minor leagues. I'd much rather have the ability to recreate both kinds.

What you quoted is about Rookie level and Single A.
If you want to recreate earlier eras of the minor leagues those levels wouldn't even exist right?
I personally don't like 30 year old scrubs in Double A, but could deal with that. If a 30 year old can't even get out of Rookie ball or Single A he should be retired... he'd be off playing fast pitch recreation leagues somewhere.

Either way doesn't really matter to me, I just kick them off my team and they go hang out on AI rosters somewhere.

[BIG Irish]

This game seems to love VORP. Use VORP. If the FA is in the top 10% in VORP than they are Type A. If the player is in the top 11-20% they are a type B. In the top 21-30% they are type C. These percentages aren't hard numbers but I think they work. The formula works wether you have a league where the MVPs have 100 VORP, or VORP of 30.

I don't really like the idea of just 5 Type A, no matter how many great players are out there. In such a scenerio three of these players wouldn't be a Type A: Barry Zito, Alfonso Soriano, Carlos Lee, Jason Schmidt, Rodger Clemens, Nomar Garciaparra, Mike Mussina, Eric Gagne.

If you look at in another way. If Clemens retires, Gagne can't come back & Mussina, Schmidt & Soriano resign that means Zito, Lee, Nomar, & then who? Sheffield & Torii Hunter become Type A? I don't think a set number is a good way to go.