Debate about people vs Computers

[CuseCubFan69]

Our little debate about hustle vs lazy put a thought in my head. What if you had the best manager in MLB vs a computer and they had the same eaxct teams. You have some guy typing in the information to the computer and he says what the team should do. Absolutely no emotion to the players, just tells them where to go and where to hit/pitch. You have the same humans as the base coaches and instructors for both teams. Which team will win more games? Why?

[Rob]

Our little debate about hustle vs lazy put a thought in my head. What if you had the best manager in MLB vs a computer and they had the same eaxct teams. You have some guy typing in the information to the computer and he says what the team should do. Absolutely no emotion to the players, just tells them where to go and where to hit/pitch. You have the same humans as the base coaches and instructors for both teams. Which team will win more games? Why?

 

I'd take Earl Weaver over a computer anyday.

[Outshined_One]

I've always felt that the talent compromising a roster is what's critical in these situations. Give a great motivator a bunch of .230 hitters and it won't matter how much he gets out of them...the fact remains that those guys are still .230 hitters and are terrible.

 

If we're to assume that you'd give the manager and computer the same rosters, I think it really depends. Managers can play stupid hunches that somehow work out while computers could go with guys who are 99% likely to succeed and still end up failing.

 

Honestly, I think the difference would be negligible and too heavily based on human behavior and performance to produce a good answer.

[A New Era]

I'd take a computer over dusty.

 

I guess it all depends on, WHO WROTE THE PROGRAM!!!

[CubsInNC]

Our little debate about hustle vs lazy put a thought in my head. What if you had the best manager in MLB vs a computer and they had the same eaxct teams. You have some guy typing in the information to the computer and he says what the team should do. Absolutely no emotion to the players, just tells them where to go and where to hit/pitch. You have the same humans as the base coaches and instructors for both teams. Which team will win more games? Why?

 

There are 2 options for this computer simulation. If you have a "AI" computer, who learns from mistakes, etc. Something like the computers and programs IBM developed to beat chess masters. A computer like that, I think would fare much better overall in a long-term competition.

 

A computer running a program that decides what to do based on a prior manager giving the programmers input, would probably not fare too well overall.

 

If my human manager was one of the best of all time, I would say that in situation 1, the computer would probably win the majority, mabye 55-45, over a span of many years after the computer learns what to do, what works, what doesn't, etc. In situation 2, I could see the computer losing about 60+% of the time. Nothing to back up my arguements, but it's a fun topic to consider.

[biittner77]

Our little debate about hustle vs lazy put a thought in my head. What if you had the best manager in MLB vs a computer and they had the same eaxct teams. You have some guy typing in the information to the computer and he says what the team should do. Absolutely no emotion to the players, just tells them where to go and where to hit/pitch. You have the same humans as the base coaches and instructors for both teams. Which team will win more games? Why?

 

There are 2 options for this computer simulation. If you have a "AI" computer, who learns from mistakes, etc. Something like the computers and programs IBM developed to beat chess masters. A computer like that, I think would fare much better overall in a long-term competition.

 

Speaking of Kasparov vs. IBM, In the matches against Deep Blue, Kasparov tried to outsmart the computer by making some unconventional moves. It didn't work because the computer wasn't bothered by Kasparov's unexpected moves.

 

I would like to see the human vs computer to see how it would affect the human managing against the computer.

[soccer10k]

We know what Joe Morgan would say.

[Northsider]

Over the short term, a *good* coach (unlike Dusty or Dusty-alikes) would be able to make decisions based on psychology - knowing a particular player is in the right frame of mind - that is to say, who's "on", or "hot", or "in the zone" or whatever, would influence field decisions, such as whether to have a player bunt or hit, or whether to use a particular player off the bench. Some managers, Stengel coming to mind, have the ability to put a player into the zone, resulting in, if not a hit, good, strong contact. A manager who has no interaction with his club, except through a computer, would not be able to accomplish this; neither would an AI (unless the AI is sufficiently advanced as to render the hardware difference irrelevant), which is a whole different can of worms.

 

In the case of a 'sufficiently advanced AI', we would now have the problem of bias; we would have to assume, in this case, that all players *AND UMPS* have the same amount of respect for an AI as they would for a Human. Otherwise, close calls (and even not so close ones) would be very liable to go against the AI's team. Further, if the AI's players resent it, then they are less likely to perform as requested and instead do as they wish, making any sort of decisions useless. Admitted, I'm putting a bad face on for the computer; but baseball hasn't exactly been known in the past for being tolerant of 'outsiders'; see Jackie Robinson. The only fix for this would be to make the AI physically indistinguishable from humans.

 

Another point: a good coach can make .230 hitters play better (over the long haul); that's what a coach is for. A sufficiently advanced (and, perhaps, disguised) AI might be able to pull of some of the same tricks, unlike a guy behind a computer, but the AI, like a human, would not necessarily know what specific method of teaching would get through to a player. Further, also like a human, the AI would not be able to 'brute force' the player by trying to cram a dozen different ideas or techniques into his head, as thinking of too many things can mess a player up.

 

Therefore: Bad coach

[chuckywang]

I believe the main value of a manager is to motivate players to maximize their performance. This can be done by arguing with umps, keeping a happy clubhouse, instilling confidence in less experienced players, or any other number of ways. As far as game situations go, most managers will more or less do the same thing.

 

Thye will look at the splits/charts, and make a percentage decision based on them, taking into account who's available. A computer can do that just as well. As far as players who are on a "hot streak", you can program that into a computer as well by looking at the stats from the last few games.

[Tracer Bullet]

We know what Joe Morgan would say.

 

"What's a computer?"

[CubsInNC]

Our little debate about hustle vs lazy put a thought in my head. What if you had the best manager in MLB vs a computer and they had the same eaxct teams. You have some guy typing in the information to the computer and he says what the team should do. Absolutely no emotion to the players, just tells them where to go and where to hit/pitch. You have the same humans as the base coaches and instructors for both teams. Which team will win more games? Why?

 

There are 2 options for this computer simulation. If you have a "AI" computer, who learns from mistakes, etc. Something like the computers and programs IBM developed to beat chess masters. A computer like that, I think would fare much better overall in a long-term competition.

 

Speaking of Kasparov vs. IBM, In the matches against Deep Blue, Kasparov tried to outsmart the computer by making some unconventional moves. It didn't work because the computer wasn't bothered by Kasparov's unexpected moves.

 

I would like to see the human vs computer to see how it would affect the human managing against the computer.

 

Hmm. I think even Deep Blue would be bothered by some of the past 4 years of managerial moves.

[Jehrico]

I'd take a computer over dusty.

 

I guess it all depends on, WHO WROTE THE PROGRAM!!!

 

How about this question...who would win between Dusty, and a computer whose running a program written by Dusty....

 

I see that game coming down as a 37 inning draw, both try to outsmart each other, both use all 25 men on the roster to pitch, and subsequently run out of players when the computer's Prior runs into Dusty's Lee on the basepaths, rendering both sides without enough players to continue the game.

[CuseCubFan69]

I could see Dusty trying time and time again on the computer to find a scenerio in which he could play Perez and Macias.

 

I think it was Davey Johnson that first started using computer stats for managing but I could be wrong.

[A New Era]

I'd take a computer over dusty.

 

I guess it all depends on, WHO WROTE THE PROGRAM!!!

 

How about this question...who would win between Dusty, and a computer whose running a program written by Dusty....

 

I see that game coming down as a 37 inning draw, both try to outsmart each other, both use all 25 men on the roster to pitch, and subsequently run out of players when the computer's Prior runs into Dusty's Lee on the basepaths, rendering both sides without enough players to continue the game.

 

Well they both sure would use statistics instead of "who's hot" because Dusty has proven to think like a computer sometimes too.

[A New Era]

Over the short term, a *good* coach (unlike Dusty or Dusty-alikes) would be able to make decisions based on psychology - knowing a particular player is in the right frame of mind - that is to say, who's "on", or "hot", or "in the zone" or whatever, would influence field decisions, such as whether to have a player bunt or hit, or whether to use a particular player off the bench. Some managers, Stengel coming to mind, have the ability to put a player into the zone, resulting in, if not a hit, good, strong contact. A manager who has no interaction with his club, except through a computer, would not be able to accomplish this; neither would an AI (unless the AI is sufficiently advanced as to render the hardware difference irrelevant), which is a whole different can of worms.

 

In the case of a 'sufficiently advanced AI', we would now have the problem of bias; we would have to assume, in this case, that all players *AND UMPS* have the same amount of respect for an AI as they would for a Human. Otherwise, close calls (and even not so close ones) would be very liable to go against the AI's team. Further, if the AI's players resent it, then they are less likely to perform as requested and instead do as they wish, making any sort of decisions useless. Admitted, I'm putting a bad face on for the computer; but baseball hasn't exactly been known in the past for being tolerant of 'outsiders'; see Jackie Robinson. The only fix for this would be to make the AI physically indistinguishable from humans.

 

Another point: a good coach can make .230 hitters play better (over the long haul); that's what a coach is for. A sufficiently advanced (and, perhaps, disguised) AI might be able to pull of some of the same tricks, unlike a guy behind a computer, but the AI, like a human, would not necessarily know what specific method of teaching would get through to a player. Further, also like a human, the AI would not be able to 'brute force' the player by trying to cram a dozen different ideas or techniques into his head, as thinking of too many things can mess a player up.

 

Therefore: Bad coach

 

This is the type of manager I prefer. One that goes with the hot bat rather than #'s. And my post earlier was about. Well what if someone programmed into the computer factors about "hot streaks" so the computer would go with the hot bat too. .