Unrealistic Replays

Unrealistic Replays (Full Version)

All Forums >> [Current Games From Matrix.] >> [Sports] >> PureSim Baseball

Message

jacmar -> Unrealistic Replays (6/1/2006 5:06:28 PM)
I played out a few seasons in computer mode, to see if the standings would be similar to the real life standings. The real life standings for 1955 had the Dodgers and Yankees finishing first. In the season replay, the Dodgers were edged out by Milwaukee, and the Yankees finished 23 games off the pace. In the 1919 real life standings, the Whitesox and Redlegs finished first. In the season replay for 1919, both teams finished way out and never were in contention at any time. Is there something that I should be doing to get these season replays to be more accurate?

VanScoy -> RE: Unrealistic Replays (6/1/2006 5:35:35 PM)
that is the nature of having a baseball universe versus a baseball simulator. Players are rated due to various factors and GM's may make different decisions due to other factors. For instance, a star player who batted .217 unexpected one year may be rated low by the game but played a lot in real life. And a hot rookie who comes on in August due to a real injury may be seen by the game as a good choice to start in April. Also, plays may break one way or the other and injuries may take a toll. it is the same way for OOTP. If you want a true year replicator, I suggest Diamond-Mind Baseball, which is unparalled in reproducing real-life playing times and stats, but no year to year career mode and no GM fun. I understand where you are coming from but its not just "pot luck" who wins.

tbear2b -> RE: Unrealistic Replays (6/1/2006 5:39:02 PM)
I'm not a big replay fan, so I'm sure someone else will come along who knows more than me, but here are my guesses as to what could be happening: 1) Injuries can play a HUGE part, so that could be affecting it in some ways. 2) While the real teams might have had some extra ooomph from an extraneous element, such as team spirit or something, the game doesn't really model this and as such, it could make the difference of a few games. 3) Doing one replay is usually not enough to tell if you are simulating it correctly. I would think that doing at least ten replays of the same season would produce a number of closer results than the one you have shown. I do know that others have done tests on the accuracy of replays and they seemed pretty solid. Again, I'm not a big fan of replays, so I don't know a ton, but I thought I'd throw in my 2 cents.

Amaroq -> RE: Unrealistic Replays (6/1/2006 8:10:19 PM)
jacmar - the factors tbear list play a big part, but there's another large factor: the AI and its choices. The AI looks at the in-game numbers and in-game stats, and we've all seen it do things like bat a clear #4 hitter in the leadoff position. That sort of choice alone can cost your guy fifty RBI over the course of the season - and then you start looking at bullpen usage, if it chose the wrong closer, if it traded somebody, of if a poor start to the season saw somebody benched - there become a dozen ways things could go 'wrong'. In 1919, for example, the teams may not have had ten solid 'real' pitchers, but the AI will manage the bullpen as though they did, putting in pitchers much weaker than actually threw an inning for the real team. Another big factor is the 'where people who got traded mid-season actually play' element. PureSim's 'replay' import will look at a player who played 80 games for the Yankees and 60 games for the Dodgers, and place him on the Yankees for the whole season - which costs the Dodgers 60 games of his talents. It could be even worse - hypothetically, if the Yankees and Dodgers traded shortstops mid-season - and the first played 68 games for the Yanks and 70 games for the Dodgers, while the second played 68 games for the Dodgers, and then 66 for the Yanks... you'd wind up with both playing all their games in Brooklyn. I typically think of a 'season replay' as a start to a career game, rather than an honest attempt to get that season to generate the correct results. If you really want to get "accurate" results, you'll need to do the painstaking thing some simmers are doing - using the baseball-reference.com to track the mid-season deals, and manually moving players from one team to another when they actually moved in real life. You'd also have to keep a close watch on the injury situation, and might need to adjust the lineups for teams yourself. I can't imagine spending that much time on it - and even if you did all of that, a few changes in random numbers can wind up in a fairly large swing away from the real-world results.

Beach23BoyP -> RE: Unrealistic Replays (6/2/2006 12:10:26 AM)
jacmar, Pure Sim is basically like OOTP where it is actually better to play with fantasy players. However, many people including myself prefer to use real players. (However, I play my own 12 team fantasy league using real players instead of the real leagues and teams.) There are too "variables" in these GM type baseball text sims to allow for a re-living of baseball history. If you want to play with real players with the real teams and therefore, again re-live Baseball history -- then you should no doubt play Strat-O-Matic or Diamond Mind. Those sims are made for perfectly (or at least as close as possible) re-living National League and American League history. Pure Sim leagues become a world unto themselves. You never know where you going. Years ago I wanted to re-live history and I played out every one of each teams games (I rolled and rolled dice day after day), but I gave that up a long time ago. Now I much prefer the "mystery" of PS or OOTP. I just need to finally start playing with "fantasy players". Then there is even more mystery!

jacmar -> RE: Unrealistic Replays (6/2/2006 12:15:33 AM)
OK guys...thanks for your input...

Amaroq -> RE: Unrealistic Replays (6/2/2006 12:56:28 AM)
Yeah, to add to what Beach said... A 'historical' player in PureSim becomes a fantasy player as soon as the import process is complete - and he's based on the player of the year you imported him, not the entire career arc of the real player. So, let's say I import Nolan Ryan at age 25. His 'current ability' is going to be based off of his age-25 performance, modulated a bit by the years on either side of it. The game may give him a huge potential based on his real peak and durability.. ... but if I then guarantee that he gets no playing time, locking him on the bench and never using him, the game is going to apply its 'player not getting used, gets worse even if he has great potential' code, and I can stunt his growth, exactly the same as I could for a high-potential fantasy player. Alternately, I might 'get lucky' and have him develop even better... So, you tend to get people frustrated when, six years down the line, the player does something different from real life, e.g., retires when he didn't IRL, or stays on an extra year or two. .. I personally enjoy that, because I'm thinking of it as 'an alternate telling of baseball history' rather than a 'recreation'... and that was intentional, on Shaun's part - he figures we want a game, where we don't know the outcome, rather than a 'replay', the ability to 'watch' the real world unfold a second time. That by itself might be kinda cool - to watch every game of the 1956 Yankees and see each at-bat, substitution, etc, play out as it did in real life... but no product on the market does that. ;)

Page: [1]

Forum Software © ASPPlayground.NET Advanced Edition 2.4.5 ANSI

1.390625