Baseball Game Theory: The 5-player simulation

In January, Matt Goldman of Beyond The Box Score hosted a most interesting thought experiment. It was an exercise in game theory, and the rules can be summarized like this (click here for Matt's post with the full details):

- One GM for each of the 30 teams
- Each GM picks 5 players from other teams to steal away and add to their roster
- If another GM picks the same player as you, neither of you get him and he stays on his old team
- You can't pick your own players
- If you successfully steal a player, you also inherit his contract situation (salary, years of control, etc.)

Who do you pick? The best players, and hope no one else blocks you? Role players, sacrificing value for a better guarantee of getting something? Sleepers, and hope no one else has the same idea?

I signed on as the GM for the Oakland A's, and here is what happened.

My strategy

I employed a specific strategy in making my picks. I figured that just looking at the entire MLB landscape and picking five awesome players would be a route to getting a lot of my selections blocked by other GMs.

Therefore, I added an angle -- instead of merely picking five players who could help my team, I chose guys who would not only help me but also hurt my biggest rivals. That allowed me to focus a bit more on less-obvious players whom I might actually get to keep, and if I gave up any value by picking slightly inferior players then I surely made it up by weakening the rest of my division (rather than weakening five other random teams).

To narrow things down, I added a few of my own restrictions to guide me:

- I focused on pre-arb players so as not to add payroll.
- I only looked at pitchers, outfielders, and middle infielders, so as to fill team needs.
- This is a fun dream scenario, so I'm not spending my time stocking the farm. I'm only taking players who will be in MLB in 2016.
- I want one player from each of the other four AL West teams, and then one more wild card.

My picks

Here was my thought process for each team. (I wrote these blurbs as I was making the picks, before I knew the results.)

Angels: Trout and Calhoun are too obvious. I thought about Huston Street, since it would be cool to bring him back to the Bay, but the A's already have enough expensive relievers. No, the pick here is Andrew Heaney. He's a 24-year-old lefty starter who just had a solid half-season in the bigs, and he still has six years of team control left. The A's can always use a quality cost-controlled starter, and without him the Angels rotation goes from questionable to downright sketchy. He was a top-50 prospect last year, so he's certainly not under-the-radar, but there are flashier pitching names out there and I'm banking that no one will land on this particular one.

Astros: Correa, Springer, Altuve, Keuchel, and Giles are too obvious. I'll go with another young starter who had a nice debut but didn't reach a full year of service -- Lance McCullers still has six seasons of team control left. Like with Heaney, he's good but not a high-profile household name, and between the two I'll bet I get at least one of them.

Mariners: Felix and Cruz are too obvious and are also somewhat pricey long-term -- still good values, mind you, but I'm trying not to add salary. Seager is good but also gets paid and isn't a middle infielder. I'm starting to get paranoid that Heaney and McCullers will get blocked, too, so I'm going further under the radar by buying slightly low on a bounce-back. Taijuan Walker has already had injury issues and posted an 83 ERA+ last year when he finally did play, but he's still only 23 this season and he's got five years of team control left. Heck, he was a top-10 prospect in baseball just two winters ago. I'm getting a cheap young starter, dammit.

Rangers: Okay, that's enough starting pitching. Time to mix it up and take an actual risk. Rougned Odor will likely be a popular pick, as a high-profile young second baseman; he's entering his age-22 season with five years of control left and he hit 16 homers last year. But the Rangers' other stars are expensive, and the meme potential from his name is worth the gamble. Besides, if he gets picked by someone else then maybe I can convince Matt that I was referring to his younger brother, who has the same name and is also in the Rangers' system?

Wild Card: One spot left. I looked at the Giants, but there wasn't a fit -- Bumgarner and Panik are too obvious, and I thought about taking Pence partly as a troll job but he's too expensive.

Nope, gotta take an outfielder, even if it means breaking a couple of my personal rules (looking outside of the division, or spending a few mil). I briefly toyed with testing the rest of the GMs by selecting Bryce Harper just to see if anyone else tried it, but that felt like a clown idea. I narrowed it down to Kole Calhoun and Randal Grichuk as two sluggers with defensive ability. Calhoun is more attractive because he's a lefty (a current team need) and he's on a division rival, but he also has one less year of control and breaks my pre-arb rule. Both guys seem like the kind of safe, good-but-not-superstar players whom other GMs might target, but let's not overthink this too hard.

I went with Kole Calhoun, even though I rejected him earlier in the process and my gut told me he was more likely to get blocked than Grichuk. It's worth rolling the dice on screwing over the Angels.

The results

The full table is kind of large and we're already at 1,000 words here, so click here to see the entire league's results.

My personal results:

Andrew Heaney: BLOCKED by the Cubs
Lance McCullers: Success! We got him.
Taijuan Walker: Success! We got him.
Rougned Odor: Success! We got him.
Kole Calhoun: BLOCKED by the Orioles

And looking at the other side of the ledger, the only player Oakland lost was ... No. 3 prospect Matt Olson, who was taken by the Mariners. No one else even attempted to steal Sonny Gray or any other Athletic.

And what about the guys I passed over? The biggest heartbreak for me was that no one ended up taking Harper. I coulda had him! ... is exactly what the other 29 GMs also thought. More importantly, nobody took Grichuk, so going against my gut to take Calhoun did not pan out. The rest, including who took them:

Rejected by me, and I was right: Trout (taken twice), Springer (SF), Altuve (SD), Seager (Was)
Rejected by me, coulda had him: Correa, Keuchel, Giles, Felix, Cruz, Bumgarner, Panik, Harper, Grichuk (not counting Street or Pence since I skipped them due to salary, not blocking risk)

And the clear winner? That was the Phillies, who went big like Ruben Amaro and ran the table on their five superstar picks: Stanton, Price, Scherzer, Greinke, Kershaw. With a huge media market and virtually no long-term money on the books, they added $129 million to their 2016 payroll (and plenty more over the next several years) by selecting high-priced guys whom many teams couldn't afford. It was a brilliant strategy by a team in a somewhat unique situation to employ it, and now they have arguably the best starting rotation in the history of mankind.

Conclusion

I'm happy with how things turned out. My picks were somewhat conservative, but nobody took any of the reigning Cy Young or MVP winners so I'm not the only one. Perhaps I could have gotten some better players, but in the end we added two cheap, high-upside starters and a long-term second baseman, all at the expense of our division rivals. Not everyone was so fortunate.

For Matt's full analysis of the results, check out his article. How do you think I did? We all had so much fun with this that we're going to take a second crack at it in February, so stay tuned for Round 2!