CPU Suicide Squeeze with the bases loaded??????

[tnixen]

Why does the CPU constantly do a suicide squeeze with the bases loaded??????

I lose so many games because of this.

You hardly ever see a team do a suicide squeeze with the bases loaded in real life!

Please Fix this Sony!!!

[nomo17k]

Re: CPU Suicide Squeeze with the bases loaded??????

http://www.operationsports.com/forum...e-innings.html

CPU becomes very fond of bunting in late/extra inning tie/1-run games. I've seen guys like Posey and Pujols sac bunt (successfully). Suicide squeeze is so frequent that it feel like it's guaranteed to happen with a guy on third in those situations...

CPU manager really needs to improve (also the game around bunting -- success rate, bunt hit variety, etc. -- still has a lot to be desired...).

[Councilmann_Jamm]

Re: CPU Suicide Squeeze with the bases loaded??????

Game this monring. Bottom of 9th. Bases loaded with 2 outs and the computer sac bunted successfully to tie the game

[JayD]

Re: CPU Suicide Squeeze with the bases loaded??????

Game this monring. Bottom of 9th. Bases loaded with 2 outs and the computer sac bunted successfully to tie the game

I've never seen an unsuccessful sacrifice bunt as of yet. It's getting to be a bit ridiculous.

[BegBy]

Re: CPU Suicide Squeeze with the bases loaded??????

I find that the CPU bunts like an absolute champion. To get on base, advance a runner, straight sacs or squeezes - whatever the situation calls for the CPU just does it. It doesn't matter who's bunting or pitching. Type of pitch or location...you can rest assured that they will get it done.

It's old at this point. It needs to change next year.

[Bunselpower32]

Re: CPU Suicide Squeeze with the bases loaded??????

I stand by what I said last year in that I think the reason for so many "clutch" infield hits and bunts is because the way this game actually works is instead of calculating the spin, velocity, and trajectory of the ball off the bat, the CPU runs a simple hit/out calculation and then picks one at random. So this year, instead of actually making a realistic physics game, they "influenced" it with the new spin mechanic. I thought they finally made it a real physics calculation, but the fact that spin appears to be applied almost at random and not in line with the PCI position a good amount of time just speaks more validation to my assumption. There's just too many plays that happen that appear tethered together, like the fielder HAS to make the catch.

I'll admit, this is the best they've covered it up ever. It's taking me much longer to get tired of it.

Normally people say "I hope I'm wrong" but I really hope I'm right, because if I'm wrong and this is what their best efforts toward a real ball physics engine is, these games are going to start showing their age real quick.

[sportomatic75]

Re: CPU Suicide Squeeze with the bases loaded??????

Great thread. I have very rarily seen the CPU either strike out or pop out in a bunting situation. Usually a successful sac bunt

[soamazen23]

Re: CPU Suicide Squeeze with the bases loaded??????

Yes. This is very frustrating. I lost a wild card game this way. Somehow Miguel Cabrera laid down a perfect bunt in the 9th inning to score the winning run from third. Great way to end a season 😐

[T_O_p12]

Re: CPU Suicide Squeeze with the bases loaded??????

Yep, super annoying. I have lost twice in extra innings now due to CPU suicide squeeze...

[raneman85]

Re: CPU Suicide Squeeze with the bases loaded??????

Ditto. Pretty bad when the CPU plays like the humans who do it online.

[nomo17k]

Re: CPU Suicide Squeeze with the bases loaded??????

I stand by what I said last year in that I think the reason for so many "clutch" infield hits and bunts is because the way this game actually works is instead of calculating the spin, velocity, and trajectory of the ball off the bat, the CPU runs a simple hit/out calculation and then picks one at random. So this year, instead of actually making a realistic physics game, they "influenced" it with the new spin mechanic. I thought they finally made it a real physics calculation, but the fact that spin appears to be applied almost at random and not in line with the PCI position a good amount of time just speaks more validation to my assumption. There's just too many plays that happen that appear tethered together, like the fielder HAS to make the catch.

I'll admit, this is the best they've covered it up ever. It's taking me much longer to get tired of it.

Normally people say "I hope I'm wrong" but I really hope I'm right, because if I'm wrong and this is what their best efforts toward a real ball physics engine is, these games are going to start showing their age real quick.

I think there are a few layers of issues around bunting in the Show that have existed for years, but the biggest issue (aside from generally robotic/predictable CPU manager AI) seems to me that possible bunting outcomes appear templatized (meaning, there are only several types of bunt hit balls reproduced based almost purely on dice roll, where as on normal swing, the hit variety is a lot greater with much finer granularity).

Not only those batted ball templates are limited (and each hit is not particularly realistic... laser bunt being ********* this year), their success rate isn't assigned (and attributed to player's bunt/drug bunt ratings) appropriately for the play to be realistic at all.


I personally think bunting should also be made based on PCI-driven, rather than almost purely based on dice roll. Rather than just pushing a button to trigger some dice roll, why should the game not make us/CPU control PCI on bunting itself?

That's not a solution to the current bunting issue itself, but I think the real issue with bunting is the lack of granular control/templates that the game currently has on those plays... and dice roll doesn't assign realistic probabilities to possible outcomes based on real-life data.

[Bunselpower32]

Re: CPU Suicide Squeeze with the bases loaded??????

I think there are a few layers of issues around bunting in the Show that have existed for years, but the biggest issue (aside from generally robotic/predictable CPU manager AI) seems to me that possible bunting outcomes appear templatized (meaning, there are only several types of bunt hit balls reproduced based almost purely on dice roll, where as on normal swing, the hit variety is a lot greater with much finer granularity).

Not only those batted ball templates are limited (and each hit is not particularly realistic... laser bunt being ********* this year), their success rate isn't assigned (and attributed to player's bunt/drug bunt ratings) appropriately for the play to be realistic at all.


I personally think bunting should also be made based on PCI-driven, rather than almost purely based on dice roll. Rather than just pushing a button to trigger some dice roll, why should the game not make us/CPU control PCI on bunting itself?

That's not a solution to the current bunting issue itself, but I think the real issue with bunting is the lack of granular control/templates that the game currently has on those plays... and dice roll doesn't assign realistic probabilities to possible outcomes based on real-life data.

Exactly, which is what I mean by "they're hiding it well this year". The hitting just has more outcomes, but I still don't think they've switched to actual physics: a free ball being hit and deflected by a real bat and fielders and what not. It's still a dice roll, and that is just not OK. It just happens to show itself more on bunting. Not only that, the increase in difficulties I'm pretty sure correspond to the AI getting a better chance on those dice rolls.

Wow, I that actually makes so much sense what you just said.

Seriously though, if anyone ever makes that game that is physics based, you better believe I'm going to get it.