Yes, another HAL story

Playoff finals. I'm down 3 games to 1 but have Scherzer in game 5 at home. Cy Young winner for the league, up 4-2,5th inning, gives up a lead off walk (still a F9 and with "hook slow" checked) and BAM replaced by Hunter as a middle reliever, next batter - HR, tie game.

OK, game goes to 9th and further. I have Perkins, Rolaids relief pitcher of the year for the league, "tied late" and "9th plus" both checked for Perkins, no sign of him, and my LHP specialist with now with a f0 is still in after 11, top of 12th other team gets 3 runs and we lose.

We all have "funny" HAL" stories, stealing home, etc., but really?? This is just not working.

JT

“So dry your tears. Fortune has not yet turned her hatred against all your blessings. The storm has not yet broken upon you with too much violence. Your anchors are holding firm and they permit you both comfort in the present, and hope in the future.”
― Boethius, The Consolation of Philosophy

“The greatest misery in adverse fortune is once to have been happy.”
― Boethius, The Consolation of Philosophy

Boethius wrote many things about fortune/luck before having the bad luck of being executed for heresy. One of the wisest was how we often forget the same luck now vexing us in a situation previously blessed us in a similar one. This is definitely the case with many SOM players. When their teams are doing well, they forget that HAL's byzantine matrix of processes/determinations behind its actions significantly contributed to our success. However, that same success makes them irrationally bemoan HAL's actions when they contribute to their failure.

As experienced SOM players, we know what we're getting when we sign up for HAL. We are getting a pretty smart program factoring in numerous inputs to make decisions most satisfying all of them. That processing of input will inevitably lead to rational actions, questionable actions favoring us, and/or questionable actions not favoring us. So, when Hal does the latter, it's not a case of HAL "not working;" it's a case of HAL working fine...just against our favor.

The moment HAL becomes a perfect program--whatever that would be--we can complain about it not working. Until then, such complaints don't recognize the known nature of HAL or the known nature of luck.

I on the other hand understand your pain.

New season coming and better Boethius to you then bb.

I neither said I didn't understand his pain, nor said feeling that pain was wrong. As I noted, HAL makes many questionable decisions; I've played SOM a long time and suffered through many of them. I said seeing HAL as "not working" because of them isn't rational, and it's not.

The fact you, and almost all of us, still look forward to a new SOM season means most of us know that. We wouldn't continue paying for a product we thought truly wasn't working, despite how often it vexes us.

Last night I WAS ranting because I was pissed. I am calm now and of course I'll keep playing.

On a unemotional note there is some point if you have to wonder if certain components (program logic flaws) of HAL are simply broken as opposed to HAL simply doing something very stupid and unpredictable, like deciding to steal home in the 9th. In real life baseball this can happen: 3rd base coach or player misconstrues a sign, player just does it on a hunch, etc.

However with such a clear directive to HAL "Slow hook" checked and the pitcher is yanked in the 5th with a F9 rating or not use a pitcher when "use tied late and use tied 9th plus" options are checked and a very tired LH specialist is kept in the game, I begin to wonder if it is a program bug.

you have to wonder if certain components (program logic flaws) of HAL are simply broken ... I begin to wonder if it is a program bug.

Broken in my opinion. The best AI algorithms account for as many combinations of scenarios and settings as possible and then defines a hierarchy of decision points for the combinations that are missed. This is one area I know really well from developing AI for turn based games in my past. From many of the results here and that I've seen in the CDROM game, the HAL code is thin.

.

Broken in my opinion. The best AI algorithms account for as many combinations of scenarios and settings as possible and then defines a hierarchy of decision points for the combinations that are missed. This is one area I know really well from developing AI for turn based games in my past. From many of the results here and that I've seen in the CDROM game, the HAL code is thin.

"Broken" implies HAL is no longer functioning as well as it once was. Considering it has been making its questionable "decisions" since the beginning, that is most likely not the case. Regardless of how HAL makes its decisions, it has still made them successfully enough to make its questionable ones relatively anomalous. As I noted above, those who rage against those actions simply forget all the times it makes rational ones helping (or even not helping) them. The fact they keep paying for SOM and HAL affirms this. If someone truly thinks HAL's code is "thin," it doesn't make sense for them to keep paying for it.

However with such a clear directive to HAL "Slow hook" checked and the pitcher is yanked in the 5th with a F9 rating or not use a pitcher when "use tied late and use tied 9th plus" options are checked and a very tired LH specialist is kept in the game

One of my favorites. The bullpen settings look clear enough, but they are unpredictable. There's a big group of managers who are in the "super-reliever" camp and lot in the "deadball era Starter" camp for just that reason. I'm in the deadball era SP camp and set the pitcher strategy to limit RP innings to as few as possible and try to get the SPs to 350 innings each. I can control RP better in the CDROM game with the Super Hal settings, but in my experience the level of control seems to be less in the online version of HAL.

I just learned another quirk (not bug) last night in a playoff game. I had my starting catcher set to "rem for D sub" checked. My backup catcher (the D dude) got hurt but HAL still replaced the starting catcher (a 4 (-1) t6) with a LF, catcher rating 5 (+5) T1-20, with a lead in the 7th. Funny thing : in a key moment Ellsbury (on my opponent's team) gets caught stealing 1 - 19 chance.