I've been thinking about Winter Baseball in the IBC. Something to fill the time until we start the draft. Caveats are it needs to be simple and needs to be wrapped up before the holidays.
Here's what I'm thinking:
1. Set up a draft in the same order we'll be drafting on Jan. 1 and let the computer draft a 40 man roster for each team. Theoretically this should level the talent, give everyone plenty of players, and put the onus on GMing up a 25 man roster . Anyone want to know what JB can do without Pujols, Mauer, Mo, Felix, Halladay, etc or what some of the rebuilders can do with a competitive roster?
2. Sim it either a week per day or a series per day. The first would be about a 24 day season and the second would be about a 50 day season. DB's every day.
3. No injuries.
Jp says we need at least half the league involved to make it viable. I'm willing to do the simming. DB can be sent out early AM, noon, or before I go to bed. Depending on what works best for people.
I was thinking maybe first pick in the 6th round as a prize for the winner.
Okay. Ideas? Yea or nay?
Winter League
-
Rockies
- Posts: 2688
- Joined: Mon Feb 06, 2012 1:00 am
- Location: Denver, CO
- Name: Nate Hunter
- Contact:
Let's do it.. But question. If a computer is going to draft for us, are we certain that we will have a complete team? I don't want, forum example, 17 pitchers and 8 position players which turn out to be all outfielders. Just curious how that would work given an autodraflt which i assume would go off BPA.
We'd be drafting 40 players per team so hopefully that won't be a problem.
This is what DMB says about computer drafting:
How the computer GM chooses players
When evaluating the available free agents, the computer GM looks at the needs of the team to make sure that it fills all of the roles (starting pitchers, relievers, players rated at each defensive position, and so on). And it looks at the ability of each player to help the team win in ways that go beyond hitting and pitching to include things like defense (range, error rates, throwing, holding runners), speed and bunting.
All hitters and pitchers are judged on a park- and league-neutral basis so the true ability of the player comes through. This is appropriate because Diamond Mind Baseball makes these adjustments when you play the games, and it doesn't make sense to ignore them when choosing talent in the first place.
If, for instance, a pitcher was effective despite facing the DH and/or playing in a hitters park in real life, he may well be chosen higher in the draft than someone whose stats look better but who had a more favorable real-life environment in which to do his work.
And you might see a slugging first baseman drop lower in the draft than you might expect. Why? Because first basemen are expected to produce more offense than players at other positions, so it's harder to stand out. And because some first basemen don't supply the other things -- defense, speed, bunting, the ability to play more than one position -- that make a player more valuable to a team.
We believe we have come up with a pretty comprehensive way to evaluate the overall contribution that these players make to a team, and for that reason, we think the computer GM will provide a worthy companion in your drafts, whether you're trying to out-think it or whether you're asking for its help in choosing players for your team.
As with any system that attempts to emulate the process that human GMs go through in something as complex as building a baseball team, there's more we can do in the future to make the computer GM even more sophisticated.
At present, it has no way to know which players are considered to be future stars despite mediocre-to-awful stats. It relies only on the statistics and ratings in the database. Since most of our Season Disks are based on a single season, that means the computer GM will consider only the player's performance in that season when making its picks.
This is what DMB says about computer drafting:
How the computer GM chooses players
When evaluating the available free agents, the computer GM looks at the needs of the team to make sure that it fills all of the roles (starting pitchers, relievers, players rated at each defensive position, and so on). And it looks at the ability of each player to help the team win in ways that go beyond hitting and pitching to include things like defense (range, error rates, throwing, holding runners), speed and bunting.
All hitters and pitchers are judged on a park- and league-neutral basis so the true ability of the player comes through. This is appropriate because Diamond Mind Baseball makes these adjustments when you play the games, and it doesn't make sense to ignore them when choosing talent in the first place.
If, for instance, a pitcher was effective despite facing the DH and/or playing in a hitters park in real life, he may well be chosen higher in the draft than someone whose stats look better but who had a more favorable real-life environment in which to do his work.
And you might see a slugging first baseman drop lower in the draft than you might expect. Why? Because first basemen are expected to produce more offense than players at other positions, so it's harder to stand out. And because some first basemen don't supply the other things -- defense, speed, bunting, the ability to play more than one position -- that make a player more valuable to a team.
We believe we have come up with a pretty comprehensive way to evaluate the overall contribution that these players make to a team, and for that reason, we think the computer GM will provide a worthy companion in your drafts, whether you're trying to out-think it or whether you're asking for its help in choosing players for your team.
As with any system that attempts to emulate the process that human GMs go through in something as complex as building a baseball team, there's more we can do in the future to make the computer GM even more sophisticated.
At present, it has no way to know which players are considered to be future stars despite mediocre-to-awful stats. It relies only on the statistics and ratings in the database. Since most of our Season Disks are based on a single season, that means the computer GM will consider only the player's performance in that season when making its picks.
-
Rangers
- Site Admin
- Posts: 4140
- Joined: Wed Feb 23, 2005 1:00 am
- Location: Prosper, TX
- Name: Brett Perryman
So I'm sure you guys who have done this before already knew this, but the draft is really terrible. I picked Dustin Ackley (758 OPS, Av d) 2nd over Buster Posey, Adam Wainwright, Ryan Zimmerman and Prince Fielder, just for example.
My highest OPS was 85th best, and that's on a 1B/DH (Adam Lind), and my best ERA and WHIP were 73rd and 66th. Those obviously don't tell the whole story, but yeah.
My highest OPS was 85th best, and that's on a 1B/DH (Adam Lind), and my best ERA and WHIP were 73rd and 66th. Those obviously don't tell the whole story, but yeah.
