Broken Bat - Suggestions: Tie Breakers

Forum >> Suggestions >> Tie Breakers

Post ID	Date & Time	Game Date	Function

#29202	01/19/2016 1:06:38 pm
MukilteoMike Joined: 08/09/2014 Posts: 3294 Inactive	I understand the problem with games for three-way ties. I doubt that happens more than once every few seasons, but whatever. Head to head and then division record is far superior to RD, which is why most leagues use those first. Again, that's about winning instead of winning by how much. The Real World Order.

#29210	01/19/2016 2:15:16 pm
newtman Joined: 11/02/2013 Posts: 3343 Inactive	Except the MLB is fall from infallible, and quite honestly I think it is a dumb method there. I have never watched an MLB wild card game, nor will I ever, as far as I'm concerned it is the dumbest move in the history of the MLB (and that is saying a lot). RD is not artificial at all, and doesn't actually favor offense as much as it seems, let's look at an example: Team A: RS: 750 RA: 700 RD: +50 Team B: RS: 690 RA: 600 RD: +90 In this case the team with the better pitching/defense won the tiebreaker. Really it favors the better overall team, which is the team that should promote anyways.

#29211	01/19/2016 2:25:57 pm
Rock777 Joined: 09/21/2014 Posts: 9821 Haverhill Halflings Legends	+1 for RD

#29233	01/19/2016 9:27:37 pm
MukilteoMike Joined: 08/09/2014 Posts: 3294 Inactive	To be fair, all tie-breakers that don't involve playing an extra game(s) are artificial. I happen to think RD is the most contrived of all, though, because it doesn't matter at all how much you win by. A win is a win is a win, and that's how they should be looked at. If we do evaluate wins, though, I think RD often rewards the wrong team. 10-1 vs. 2-1. Which win is better? I would argue that it's the 2-1 victory. Again, there are many reasons why RD is inferior. The real world agrees in every sport except soccer. I think there are valid reasons for it for soccer, but I won't digress into that discussion. Oh, by the way, if you're not watching any wild card games, you're missing out. They are intense.

#29235	01/19/2016 11:56:39 pm
MukilteoMike Joined: 08/09/2014 Posts: 3294 Inactive	I did a bit of quick research. From 2025 to 2020, the number of two-team ties for the division lead were 5, 3, 4, 3, 2 and 4. I was surprised it wasn't more than that (2.8%). There was only one three-way tie. That one is fascinating. Check out league V.15 in 2022. Before looking, guess how many wins they had and the best RD. I'll pause here and give you a shot. Their records were seventy-five and eighty-five. Yep. With the winner having a minus forty-six RD. Yep yep. Crazy, huh?

#29263	01/20/2016 4:18:04 pm
Rock777 Joined: 09/21/2014 Posts: 9821 Haverhill Halflings Legends	While I understand your frustration here are a few reasons why I think RD works particularly well in BB. 1) Everyone in your division plays exactly the same number of games against exactly the same opponents. 2) You can't decide to take it easy after you have a one run lead. You also can't decide to run up the score. The game is played out with the same engine strength no matter what. Because this is not the real world, it works much better for the game then it would in real life. I couldn't make my pitcher's throw table tennis matches if I wanted to. 3) Why would a 2-1 victory ever be better than a 10-1 victory? This is a tie breaker, so we've already proven that both teams can win the same number of games. How else could you possibly measure the difference? To say that RD doesn't predict the better team is to say that you don't believe in the Pythagorean Theorem. Granted its only a Theorem, but there is an awful lot of evidence that its pretty good. EDIT: Head to head is a small sample size. Not really a good predictor. vs Division is extremely arbitrary. Why should division wins count for more? 50% chance the other division was better, so its more impressive to have the wins over there. Both teams played the other division teams exactly the same number of times. Updated Wednesday, January 20 2016 @ 4:24:04 pm PST

#29264	01/20/2016 4:32:53 pm
Spoonerific Joined: 01/17/2013 Posts: 339 Inactive	The only argument I have for RD vs. record is that a team that averages fewer runs but gives up fewer is at a disadvantage. Take the 2-1 example. If one team averaged a 2-1 score they would lose the tiebreaker to a team that averaged a 3-1.99 average. Which is more impressive? I'd be fine with P-T as the first tie-breaker as it takes that bias out of the equation [so to speak]. I'm not sure how easily that can be implemented as the breaker though.

#29267	01/20/2016 5:26:08 pm
newtman Joined: 11/02/2013 Posts: 3343 Inactive	Spoon, your example is arbitrary and not necessarily true. If one team averaged 3-1 and another averaged 5-4 the team that scored and gave up fewer would win.

#29268	01/20/2016 5:48:55 pm
MukilteoMike Joined: 08/09/2014 Posts: 3294 Inactive	I understand your frustration here are a few reasons why I think RD works particularly well in BB. I've talked about this a few times before. This is the first time it has impacted me. I would have been saying the exact same thing even if I had the RD advantage. In fact, I would probably hate promoting that way even more than I hate losing this way. You can't decide to take it easy after you have a one run lead. You also can't decide to run up the score. This isn't true. Who is the most important person on the field? Obviously it's the pitcher by a large margin. If you're getting destroyed in a game, your worst pitchers will come in. The team in the lead uses better pitchers because they're in the lead. Why would a 2-1 victory ever be better than a 10-1 victory? For the same reason I mentioned above. In a close game, the best pitchers are used by both teams. There are more late inning moves and strategies that only get implemented in close games. Both teams are doing all they can in these to scratch out a win. This is a tie breaker, so we've already proven that both teams can win the same number of games. How else could you possibly measure the difference? Easy. Play another game. They won't have the same number of wins then. MLB agrees. To say that RD doesn't predict the better team is to say that you don't believe in the Pythagorean Theorem. Granted its only a Theorem, but there is an awful lot of evidence that its pretty good. It is brilliant for triangles, while only being fun for baseball due to the reasons I've stated, plus some that haven't been discussed. Head to head is a small sample size. Not really a good predictor. vs Division is extremely arbitrary. Why should division wins count for more? Finally someone brought this up. I didn't bring it up because I don't think everyone considers this properly and I hate giving away 'secrets' and advantages. Head to head and division games are more important because those are the teams you are directly battling for the crown. Every win for you is a loss for your direct competitors. When you play the other division, the win is only for you; your divisional opponents can still win their game. That means head to head and divisional games are twice as important as cross divisional games. That's the biggest reason every major American league with the exception of MLS uses those tie breakers ahead of scoring differential. Updated Wednesday, January 20 2016 @ 5:58:18 pm PST

#29270	01/20/2016 6:37:15 pm
Rock777 Joined: 09/21/2014 Posts: 9821 Haverhill Halflings Legends	I've talked about this a few times before. Yep, I wasn't questioning your motives, I know this has been an issue for you since long before it ever impacted you. This isn't true. Who is the most important person on the field? Obviously it's the pitcher by a large margin. If you're getting destroyed in a game, your worst pitchers will come in. The team in the lead uses better pitchers because they're in the lead. Accept that is all determined by the engine based on your game settings. My point is that you can't "run up the score" dynamically in game. Your settings at set before the game, and everyone plays with the same engine. For the same reason I mentioned above. In a close game, the best pitchers are used by both teams. There are more late inning moves and strategies that only get implemented in close games. Both teams are doing all they can in these to scratch out a win. I'd disagree here as well. This is looking at the game as being decided by pitching. If I have better batting and score 10 runs, why does that make me a lesser team? Finally someone brought this up. . . . That means head to head and divisional games are twice as important as cross divisional games. That's the biggest reason every major American league with the exception of MLS uses those tie breakers ahead of scoring differential. This is flawed thinking. Every game has equal importance when determining how good a team is. Using your logic you would actually be double counting, because you get extra credit for winning in the division to start with, and then you use it as a tie breaker on top of it... If team A beats team B, and team B beats team C, and team C beats team A, who is the better team? This happens all the time, and is part of the reason college football rankings are totally bunk. RD shows an aggregate. Head to head show "on any given Sunday". Statistically the aggregate is going to be a better predictor of "goodness" than any small sample size of data. Anyone can get lucky over a 5 game stretch. Getting lucky over a 160 game stretch is a lot harder. That's the whole premise of Monte Carlo simulations. If you ran your Monte Carlo a total of 5 or 10 times, your results wouldn't be considered valid. 160 times and they probably would have some validity (depending on model complexity).