SIM UPSET MODE - Page 2 - Hoops Dynasty

Posts: 5030 (0)

Hall of Famer

95

Veteran

Posted by supersloth33 on 5/6/2015 8:35:00 AM (view original):
I seem to lose a couple games about every season to sims I should have no right losing to. A couple seasons ago I lost to a couple teams I was favored by 40+. However, I don't recall winning games the past 3 or 4 seasons that I shouldn't. I've learned to roll with the punches, hoping I'll get the wins I'm not supposed to get sometime soon, preferably in the NT.

However, I am getting really sick of people saying, "It happens in real life too", like that's the be all end all excuse for anything that can't be explained away. But when you bring up a suggestion like whitey34 made in another topic about giving teams home court advantage in the D3 NT early rounds games, it gets no response from those of you whom like to point to real life. Really frustrating when peoples' philosophies are so inconsistent. Sorry, just felt the need to vent.

i think the "it happens in real life" is kind of weak as well. here, that is not really the main point though. the main point is that upsets happen in general (in HD) and with sims that is no different. even games you win 99% of the time, you lose 1%, losses to crap sims are inevitable over a long enough time period.

if someone is losing to sims more than they "should", then its probably an issue with their team or their game plan.

5/6/2015 12:39 PM

dahsdebater

Posts: 10036 (0)

Hall of Famer

284 270 18

Pro

I've never understood the conspiracy theorists on sites like this. In the end, it's illogical, it doesn't make sense. The people running the site want your business. They're not going to include ghost code in the sim engine just to screw with you. That would be horrible business. The only reason they'd include an "upset mode," as you call it, would be if they perceived it as a feature. It seems unlikely that causing the paying customers to lose to teams nobody's paying for would be considered good business, but I guess you could envision a scenario in which somebody thought that leveling the playing field a little more would be a good idea. But if they were going to do something like that, again, they'd have to see it as a positive in the game. And they would absolutely announce it. They're not going to add features and NOT advertise them to the customer base.

There's just no logical reason to include something that involves extra coding complexity and not tell the users about it. People get upset at bad losses and I guess don't think things through, but it really doesn't make any sense. A few people in the SLB community are convinced that a few players will be "poisoned" or "bumped" at the beginning of a season, or at random intervals. Same logic should apply - if somebody took the time to program that in, they would advertise it as a feature. I think, in the end, it all stems from a poor understanding of statistical variance, sample size, an possibly team/game planning. Although grizzly_one was one heck of a SLB player.

5/6/2015 12:55 PM

therewas47

Posts: 657 (0)

Hall of Famer

127 1

Prospect

I havent looked at the users assuming conspiracy but maybe you all are not that good. I get just as mad as anyone over bad breaks but if your losing games against sims on a even simi-consistent basis then your teams are probably not all that good.

5/6/2015 1:38 PM

supersloth33

Posts: 166 (0)

Hall of Famer

47 7

Prospect

I think your logic is a little flawed as well, dahsdebater. If this site was truly run to want our business, things wouldn't be how they are. There are numerous things in the coding that is not publically acknowledged. For example, it's a generally accepted concept that the fullcourt press is too dominant. They have never admitted that to anyone that I am aware of, but it's there. There are other examples that have been discussed on the forums that also highlight this point. I'm not saying that there is a code in the game that specifically allows for off games and upsets, but I'm also saying you can't point to your logic and say for sure something is not in there. I know I haven't gotten a peak at the code and I'm pretty sure you haven't either. To declare with absolute certainty anything is a mistake.

As for therewas47, I don't know if you were talking to me specifically, but the two games I spoke of I was a 69 point and 48 point favorite based on the logic of this gaming system. So, that was saying I was a pretty superior team to the team I was playing. Given that logic, if I were to lose, it's by definition an upset, not that my team wasn't as good as I thought it was. The game itself said I was that much better than the team I was playing. If I allow for your point that my team just wasn't that good, then you must admit there is something wrong with the logic in the game itself when it comes to matchups. You don't get it both ways. Either it's an upset or the logic itself is flawed.

5/6/2015 5:35 PM

davis

Posts: 985 (0)

Hall of Famer

639 2 6 324 21 31

Pro

This is pretty simple... if your team is so good that it should have a 90% chance of winning each individual game, then your expectation is that you'd lose once out of every ten games. In a 30 game season, you'd expect 27 wins and (based on the binomial distribution) could reasonably land anywhere from 24 to 30 wins. The math isn't difficult.

5/6/2015 6:57 PM

dahsdebater

Posts: 10036 (0)

Hall of Famer

284 270 18

Pro

I hate to say this, because I didn't like the way therewas47 called people out, but you're kind of exactly proving his point. Betting lines a) ignore gameplanning, so "assume" you're running your team a certain way; you could be coaching worse than that and b) are generally known by everyone to be fairly poor. And your example of "not publicly acknowledged (sic)" coding issues is very misplaced. They don't code any given defense to be "strong." You can, if you want, think about defenses similarly to character classes in RPGs, if that helps - each one is designed to have strengths and weaknesses. When the game is first released, they aren't always well balanced, and sometimes a strength or weakness needs to be toned up or down for one or more classes. One might argue that in this game balancing hasn't been handled very well. I actually don't think press is a lot stronger than man, personally, but that's somewhat beside the point. The point is that they don't intend for one defense to be stronger than the other. If it happened that way, it was due to a lack of effort. A lack of effort is a common theme in analyzing this game. There is a very big difference between a lack of effort in balancing defenses and putting in extra effort to randomly throw certain games. A massive difference.

5/6/2015 7:57 PM

possumfiend

Posts: 511 (0)

Hall of Famer

130 5

Pro

Posted by davis on 5/6/2015 6:57:00 PM (view original):
This is pretty simple... if your team is so good that it should have a 90% chance of winning each individual game, then your expectation is that you'd lose once out of every ten games. In a 30 game season, you'd expect 27 wins and (based on the binomial distribution) could reasonably land anywhere from 24 to 30 wins. The math isn't difficult.

Actually it's not that simple. Your chance of winning each individual game is not static and is dependent upon the quality of opponent you play in any given game. Log5 methodology (credit to Bill James) is often utilized to predict a team's likelihood of winning based upon the "true winning percentage" of the teams playing. So a team with a 90% chance of winning a game against an "average" team (one with an exactly 50% chance of winning against another "average" team) would be expected to win against that particular team 90% of the time. However the same team would be expected to win just 50% of the time if opponent also had a 90% chance of winning against an "average" team; and they would be expected to win 98.8% of the time against teams with a 10% chance of winning against an "average" team. So really expectations of winning any particular game is dependent upon the team actually being played and expectations on a certain number of wins in a season is actually dependent upon analysis of each game in a team's schedule.

I guess my point is that if your team is good enough to win 90% of the time against an average team, you should be expecting to win more than 90% of the time against below average SIM teams.

5/6/2015 8:36 PM

davis

Posts: 985 (0)

Hall of Famer

639 2 6 324 21 31

Pro

Possum, I understand that, but the point still stands: stuff happens when randomness is involved. If the choice is between probability theory and conspiracy theory, I'll go with probability every time. The notion of "upset mode" is ridiculous.

5/6/2015 9:20 PM

possumfiend

Posts: 511 (0)

Hall of Famer

130 5

Pro

Understood and agreed. I just felt the need to correct because there's a lot of bad "probability theory" out there too.

5/6/2015 9:27 PM

davis

Posts: 985 (0)

Hall of Famer

639 2 6 324 21 31

Pro

I guess the way to really answer the question of expected wins and win variance in a given season would be to compute the expected value and variance of a sum of Bernoulli random variables, one for each game, where p varies from game to game based on quality of opponent. The hard part, of course, is figuring out reasonable values of p for each game.

Either way, most long-time coaches have reconciled themselves to the fact that there will be one or two inexplicable results per season when the die rolls just don't go your way. I go into each season with expectations for the season-long performance of my squad, but I go into individual games knowing that goofy stuff can take place.

5/6/2015 9:41 PM

nachopuzzle

Posts: 1427 (0)

Hall of Famer

105 2

Prospect

Posted by Trentonjoe on 5/5/2015 10:11:00 AM (view original):
If something has a 1% chance of happening, then after a 1,000 career games it should happen, oh ten times!

Weird crap happens. Wait until the DEF player of the year fouls a guy with 0:01 seconds on the clock and the D FT shooter sinks two shots to win in the NT game. Tho

...are you relying on the Anchorman for your info about statistics, because I'm pretty sure those were just jokes dude?

5/7/2015 12:06 AM

emy1013

Posts: 2731 (0)

Hall of Famer

828

Prospect

Weird games DO happen and when we're on the wrong end of one, we certainly remember it much clearer than when we're on the good end. A few people have heard this example, so to those, please bear with me. I was a 62 point favorite in a game against a pretty bad conference Sim team. He had centers and power forwards draining threes and no matter who was guarding them, they just wouldn't miss. I ended up losing in OT by 2 points. For the other 34 games that season, I lost 1 time. And that 1 time was not in the National Title game. Yes, a 33-2 National Champion had a 2 point regular season loss to a crappy Sim team that they were favored to beat by 62 points. It happens. It sucks. It ain't going away.

5/7/2015 12:39 AM

llamanunts

Posts: 1498 (0)

Hall of Famer

29 95

Hall of Famer

The non-upset-mode people are right about this. To sum up:

1. There is NO INCENTIVE for WIS to have programmed in some kind of upset mode.

There is nothing for them to gain by doing that. And if, for some reason, WIS wanted huge upsets to happen every now and then...

2. There is NO NEED to do any extra coding.

As noted repeatedly, unexpected results are going to happen within small samples of randomized outcomes, even weighted ones, and that's what each game literally is: a series of weighted, randomized outcomes. If they WANTED to make sure these types of games happened, there is still nothing extra they needed to do to make them happen.

5/7/2015 11:31 AM

gillispie1

Posts: 5030 (0)

Hall of Famer

95

Veteran

Posted by possumfiend on 5/6/2015 8:36:00 PM (view original):
Posted by davis on 5/6/2015 6:57:00 PM (view original):
This is pretty simple... if your team is so good that it should have a 90% chance of winning each individual game, then your expectation is that you'd lose once out of every ten games. In a 30 game season, you'd expect 27 wins and (based on the binomial distribution) could reasonably land anywhere from 24 to 30 wins. The math isn't difficult.

Actually it's not that simple. Your chance of winning each individual game is not static and is dependent upon the quality of opponent you play in any given game. Log5 methodology (credit to Bill James) is often utilized to predict a team's likelihood of winning based upon the "true winning percentage" of the teams playing. So a team with a 90% chance of winning a game against an "average" team (one with an exactly 50% chance of winning against another "average" team) would be expected to win against that particular team 90% of the time. However the same team would be expected to win just 50% of the time if opponent also had a 90% chance of winning against an "average" team; and they would be expected to win 98.8% of the time against teams with a 10% chance of winning against an "average" team. So really expectations of winning any particular game is dependent upon the team actually being played and expectations on a certain number of wins in a season is actually dependent upon analysis of each game in a team's schedule.

I guess my point is that if your team is good enough to win 90% of the time against an average team, you should be expecting to win more than 90% of the time against below average SIM teams.

he was over simplifying but that is plenty good enough of an approximation for the purposes of this discussion.

5/7/2015 11:56 AM

gillispie1

Posts: 5030 (0)

Hall of Famer

95

Veteran

Posted by supersloth33 on 5/6/2015 5:36:00 PM (view original):
I think your logic is a little flawed as well, dahsdebater. If this site was truly run to want our business, things wouldn't be how they are. There are numerous things in the coding that is not publically acknowledged. For example, it's a generally accepted concept that the fullcourt press is too dominant. They have never admitted that to anyone that I am aware of, but it's there. There are other examples that have been discussed on the forums that also highlight this point. I'm not saying that there is a code in the game that specifically allows for off games and upsets, but I'm also saying you can't point to your logic and say for sure something is not in there. I know I haven't gotten a peak at the code and I'm pretty sure you haven't either. To declare with absolute certainty anything is a mistake.

As for therewas47, I don't know if you were talking to me specifically, but the two games I spoke of I was a 69 point and 48 point favorite based on the logic of this gaming system. So, that was saying I was a pretty superior team to the team I was playing. Given that logic, if I were to lose, it's by definition an upset, not that my team wasn't as good as I thought it was. The game itself said I was that much better than the team I was playing. If I allow for your point that my team just wasn't that good, then you must admit there is something wrong with the logic in the game itself when it comes to matchups. You don't get it both ways. Either it's an upset or the logic itself is flawed.

dahs is right, your example is nothing like his. we know there is code for the press. just because it doesn't work exactly how we might like, we know its there. its apples to oranges.

dahs makes a fair point, its a reasonable line of thinking. it doesn't prove anything but generally speaking he is right. it makes no sense for WIS to put in a specific "lose to sims" function, that would be absolutely ridiculous from a business perspective. its safe to assume from that line of thinking, it doesn't exist. we also have had admin confirm there is nothing like that, and from the experience of a great many coaches, we have no reason to think such a thing exists.

also just FYI, the sim doesn't run your game ahead of time, to come up with those spreads. those spreads were added for a fun little scheme where you could bet on games. itw asn't widely used and the spreads were so ridiculous, that the system was pretty crappy. so they scrapped it. the spreads are just an artifact. they are simple piece of logic that really has nothing to do with the sim engine itself. so, thinking "the game thinks you should win" is really not a fair characterization of how it works. some crappy logic added after the fact that is known to be terrible, thinks you should win. its kind if colonels thinks someone cheated, and then you said, the userbase thinks that person cheated. it wouldn't be 100% incorrect in either case, but for practical purposes, it is.

end of the day, its pretty obvious to most people who have been here a long time, that there is no such thing. most likely the people who are struggling vs sims are either seeing a couple bad dice rolls as a trend, when there isn't one, or are not playing against the sims correctly. if you guys want to post a couple box scores, im happy to take a look and i can let you know what the situation is, is it a bad dice roll, is it your game plan, or is the difference in teams just not as big as you are thinking. its got to be one of those 3 (or some combination thereof).

5/7/2015 12:05 PM (edited)

Forums

SIM UPSET MODE Topic

Search Criteria