SubscribeShould you favor keeping better players, even if they represent a smaller savings? This issue came up in the comments on “How Do You Decide On Keepers?” and I want to spend some time addressing it.
Let’s stick with the Justin Morneau example from last time–a $20 player who can be kept at $14. And let’s pretend that we also have David Wright available at $42–the exact amount that what we have him projected at. We can either keep Morneau and save $6 or lock up Wright for no savings.
Which one of those would you keep?
Wright is definitely the better player, and our team is better with Wright than with Morneau. But this isn’t actually an either/or situation, because our final team isn’t based just on who we keep. Keeping Morneau doesn’t prevent us from drafting Wright, and keeping Wright doesn’t preclude drafting Morneau.
So we are concerned not just with how much our keepers are worth, but also with the value of players we can acquire at the draft if we kept certain players.
If we keep Morneau and draft Wright for $42, we have saved $6 overall. If we keep Wright at $42 and draft Morneau for $20, we haven’t saved anything. Keeping Morneau seems to be the better deal.
Inflation
That ignores inflation, though. Keeping players always creates inflation, and inflation has a greater effect on the top-tier players (see “Should Inflation Affect Your Keeper Choices?“). Thinking back to high school algebra, we can solve a simple equation to find the point where inflation makes keeping Wright or keeping Morneau equal propositions (i.e. the cross-over point):
$42x + $14 = $20x + $42
$22x = $28
x = 1.27
So if inflation is 27%:
We can keep Morneau at $14 and draft Wright for $53 ($42 * 1.27), paying $67 for $62 of value.
Or we can keep Wright at $42 and pay $25 for Morneau ($20 * 1.27), also paying $67 for the same $62 value.
If inflation is any amount above 27%, Wright is the better deal, even if he isn’t kept below value. But any amount below 27% favors Morneau.
The Bottom Line
I doubt most leagues have 27% inflation, except for ones that lets teams keep players for extremely long time periods at their original prices. (I’m picturing a league where Hanley Ramirez and Johan Santana are $1 keepers…) If that’s the case, then it’s unlikely that you are thinking about keeping Wright at $42.
I’m guessing that inflation in the typical keeper league is closer to 10-15%. At that rate, I think Morneau is still the easy choice.
Related posts:
You are correct that the longer you can keep the player, the more the inflation. Another huge factor is how many players you can keep. We have a long structure (salary raises $5, $5, $7, $10, $10…etc, and with redshirts) and a lot of keepers (15). Inflation was easily 75% on drafted players $15 or more last season.
Here is a little game I devised. This is a 12 team league based on last years stats. Since we won’t be trading or anything, we’ll be looking at straight $value and positions don’t matter. I just pressed “get values” on the price guide to get the $ values. Everybody will have the same gameplan of not going over value on a player…except Team 1. He’s going to bid $1 more on the best player available until he can’t afford to. Then Team 2 will catch on and do the same. Team 3 will be a little slower to the party and catch on after Team 2. Eventually Team 1 will have enough money to buy another player at +$1. When he does, he will do so. This will go on and on. The only team that is not going to overspend on a player is Team 12. They will only be able to buy a player when the other 11 teams can’t afford to buy a player. Once nobody can afford to overspend for a player, the best player goes to the max bid. Once every team has a max bid of $1, snake it. My guess is, Team 12 will end up with 23 players worth about $8 each. That’s $184 in value. The $76 they don’t get will be spread to the other teams with Team 1 gaining the most value, Team 2 second…etc.
For example, Team 1 gets;
Pujols, $36
Hanley, $35
Wright, $34
Manny, $34
Reyes, $33
Pedroia, $32
Halladay, $32
That would leave them $24 to buy 17 players, so a max bid of $8, which would eventually give them a $7 player like Casey Blake.
I don’t know about you, but even if that team filled in with waiver guys, it is a team I wouldn’t want to face in an eight team league, let alone a twelve.
Team 2 gets;
Holliday, $31
Berkman, $31
Hamilton, $31
A-Rod, $30
Beltran, $30
Mauer, $30
Sabathia, $30
RIvera, $30
That leaves them $17 to buy 15 players with a max bid of $3. They will probably end up getting a player worth more than $3 because everybody will be out of money.
You are forgetting an opportunity cost which plays into the inflation part of it.
How many players are there of Wright’s caliber, not quite half a dozen? Also to a very minor extent at his position.
How many are in that $20 range for Morneau. Something around 20, 25 of them? Many of them 1b already.
There is something to be said for going to the draft and not having to worry about Wright because you own him. If someone else comes in and wants Wright and you get outbid your plan B is a far worse option.
If you lose out on Morneau you go get Adrian Gonzalez, Derrek Lee, Travis Hafner, etc (and really Morneau as a $20 player in a mixed league?)
That reduces your threshold considerably. Probably from that 27% to around 15% (which many leagues have).
At the end of the day it is who you have on your roster. Your roster looks better with Wright and a Morneau clone than it does with Morneau and Scott Rolen.
@rwperu34: Your game is interesting, as it show what happens when a market becomes too inefficient (e.g. one team leaving large amounts of money on the table, everyone else getting players for a fraction of their value).
Dollar values (and not just mine, either) have to assume that the auction is an efficient market. For them to be accurate:
a) Everyone must spend most of their money.
b) Every player must be bought for close to their projected price.
As those assumptions are violated, the projected values become less and less meaningful.
@Mosey: I’m not sure what you mean by opportunity cost here.
However, your comparison is a false one: You make it sound like the choice is Wright + A.Gonzalez vs. Morneau + Rolen. But you leave out how that second scenario still has the $42 you didn’t spend on Wright. Even if you missed out on Wright, you are still going to spend the money elsewhere. As long as that money is spent on $42 of value, you are going to come out better by keeping Morneau.
“and really Morneau as a $20 player in a mixed league?”
Morneau’s value depends on if you think his 2009 performance will more closely resemble 2008 or 2007. A $20 projection assumes that he will be somewhere in between those two.
Under what parameters are you assuming inflation is only 10-15%? For example, in my AL-only league (12 teams, 4×4, 14-9 split, 10 keepers, 3 year contracts), I’d say inflation actually is around the cutoff in your example.
@Mays: If overpaying a little for top flight talent nets you the best team, isn’t that the most efficient way to draft?
My guess is, for all the teams to end up with equal value, you’ll have to overspend by $3-$4 on the top players. That’s 10%-15% inflation in a non keeper league that’s looking back (ie no wiaver guys turning into $5 players). Of course you could end up in a league where everybody decided they are going to pay $5 less no matter what. In that case, you get to buy for $4 less and the beating would be even more violent.
In order to insure you get all the money spent, you overpay for the guys at the top. A guy in my league year in and year out will end up with $20-$30 left on the table at the end of the draft. Every year, he says the same thing. “Instead of buying that $20 player for $20, I should have spent $40 on that $30 player.”
The biggest “problem” with our $values is, we underestimate replacement level. There is a big difference in looking backwards in October than looking forward in March. In October, if you draft that guy, you get that production. In March, that’s not true. That guy with the $1 OF will have at least a $5 player in there by the middle of May. That guy that drafted the $1 pitcher is going to get at least $10 worth of value by rotating two start pitchers, good matchup starters, and middle relievers. Or by just taking dominant starters that didn’t have enough innings. Using the performance model will make things a little closer, but still not nearly accurate.
Perhaps the solution is to adjust replacement level? If I set the replacement level higher by 2 pts, then the top player had 47% inflation. Break even was ~$20 and 38 players dropped out, losing over $100 in value.
Jurgen: I was thinking about a league with players kept for one year, mainly because that’s what my league does. :-) So maybe I shouldn’t have been as confident that most leagues wouldn’t face 27% inflation.
If your league lets teams keep players for multiple years (even with some sort of +$5 system), you are correct that it is not unrealistic to expect that kind of inflation.
@rwperu34: Overpaying only gets you the best team if it is like your example, where everyone overpays so that there are underpriced players late.
If everyone had perfect dollar values and one team overpays by $4 on everyone while the other teams stuck to the values, I have no doubt that first team would be on the bottom.
Now, it can make sense to adjust the replacement level, but I would do it for a different reason:
Near the bottom of the draft pool, teams will have more disagreement about prices. A guy you have listed for $6 might go for $1. Someone else whom you rank below replacement level might go for $2.
This disagreement frees up a little extra money, money that should probably be worked in to the inflation rate.
Good point on disagreement about draft prices, although I’d think it would have even more of an impact at the top. Not on a % basis and it’s unlikely somebody will rate a player you have below replacement at $30, but there could easily be a +/-$10 swing ($5 minimum) on every player. That’s what is going to create the “inefficient” market outlined in my little game. Without a superstar premium, at least one person in a 12 team league is going to have every superstar and star worth more than you do. That sets up a situation where you end up not overspending on anybody, but your best player is worth $20. As we know, in fantasy baseball $39+1>$20+$20.
I’ll go as far as to say, if you give a team Hanley, Reyes, Wright, A-Rod, Pujols, and $60, that team is a heavy favorite to win and finshes in the top three 90% of the time.
“If your league lets teams keep players for multiple years (even with some sort of +$5 system), you are correct that it is not unrealistic to expect that kind of inflation.”
Pfew.
For a moment, I thought one of us had gone insane. ;)
(And yes, we have some sort of +$5 system.)