Latest NBA results used: Apr 13 Western Conference Average Wins Code: Team Wins ==== ==== Lakers 64.8 Nuggets 54.3 Spurs 53.7 Blazers 53.7 Rockets 53.5 Mavericks 49.5 Hornets 49.3 Jazz 48.2 Western Conference Wins (Percent) Code: Team 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 ==== ===== ===== ===== ===== ===== ===== ===== ===== ===== ===== ===== ===== ===== ===== ===== ===== ===== ===== Lakers 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 23.0 77.0 Nuggets 0.0 0.0 0.0 0.0 0.0 0.0 66.1 33.9 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Spurs 0.0 0.0 0.0 0.0 0.0 31.6 68.4 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Blazers 0.0 0.0 0.0 0.0 0.0 33.9 66.1 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Rockets 0.0 0.0 0.0 0.0 0.0 53.2 46.8 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Mavericks 0.0 46.8 53.2 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Hornets 0.0 68.4 31.6 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Jazz 77.0 23.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Western Conference Placement (Percent) Code: Team 1st 2nd 3rd 4th 5th 6th 7th 8th out ==== ===== ===== ===== ===== ===== ===== ===== ===== ===== Lakers 100.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Nuggets 0.0 69.3 30.7 0.0 0.0 0.0 0.0 0.0 0.0 Spurs 0.0 0.0 36.2 11.1 52.7 0.0 0.0 0.0 0.0 Blazers 0.0 0.0 11.4 65.4 23.2 0.0 0.0 0.0 0.0 Rockets 0.0 30.7 21.8 23.5 24.0 0.0 0.0 0.0 0.0 Mavericks 0.0 0.0 0.0 0.0 0.0 36.2 52.9 11.0 0.0 Hornets 0.0 0.0 0.0 0.0 0.0 56.3 35.5 8.2 0.0 Jazz 0.0 0.0 0.0 0.0 0.0 7.6 11.6 80.9 0.0 Western Conference Division Title (Percent) Code: Team Northwest Title Percent ==== ====================== Nuggets 100.0 Blazers 0.0 Jazz 0.0 Team Pacific Title Percent ==== ====================== Lakers 100.0 Team Southwest Title Percent ==== ====================== Spurs 36.2 Rockets 63.9 Mavericks 0.0 Hornets 0.0 Western Conference Playoff Advancement (Percent) Code: Team To 1st To 2nd To Conf To Fnls Champs ==== ======= ======= ======= ======= ======= Lakers 100.0 85.7 64.9 52.5 22.6 Nuggets 100.0 68.0 37.2 11.3 1.9 Spurs 100.0 49.1 19.0 6.7 1.1 Blazers 100.0 62.9 23.9 13.3 3.1 Rockets 100.0 59.2 28.3 10.5 2.0 Mavericks 100.0 28.8 10.3 1.8 0.2 Hornets 100.0 26.9 9.3 1.6 0.1 Jazz 100.0 19.4 7.3 2.3 0.3 Eastern Conference Average Wins Code: Team Wins ==== ==== Cavaliers 66.9 Celtics 61.6 Magic 58.8 Hawks 47.3 Heat 43.0 Bulls 41.7 76ers 40.5 Pistons 39.4 Eastern Conference Wins (Percent) Code: Team 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 ==== ===== ===== ===== ===== ===== ===== ===== ===== ===== ===== ===== ===== ===== ===== ===== ===== ===== ===== ===== ===== ===== ===== ===== ===== ===== ===== ===== ===== ===== Cavaliers 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 14.1 85.9 Celtics 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 2.0 36.7 61.3 0.0 0.0 0.0 0.0 0.0 Magic 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 17.1 82.9 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Hawks 0.0 0.0 0.0 0.0 0.0 0.0 0.0 11.9 44.8 43.3 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Heat 0.0 0.0 0.0 24.2 54.5 21.3 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Bulls 0.0 0.0 28.4 71.6 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 76ers 0.0 56.1 39.0 4.9 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Pistons 63.7 36.3 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Eastern Conference Placement (Percent) Code: Team 1st 2nd 3rd 4th 5th 6th 7th 8th out ==== ===== ===== ===== ===== ===== ===== ===== ===== ===== Cavaliers 100.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Celtics 0.0 100.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Magic 0.0 0.0 100.0 0.0 0.0 0.0 0.0 0.0 0.0 Hawks 0.0 0.0 0.0 100.0 0.0 0.0 0.0 0.0 0.0 Heat 0.0 0.0 0.0 0.0 100.0 0.0 0.0 0.0 0.0 Bulls 0.0 0.0 0.0 0.0 0.0 84.1 15.9 0.0 0.0 76ers 0.0 0.0 0.0 0.0 0.0 15.9 84.1 0.0 0.0 Pistons 0.0 0.0 0.0 0.0 0.0 0.0 0.0 100.0 0.0 Eastern Conference Division Title (Percent) Code: Team Southeast Title Percent ==== ====================== Magic 100.0 Hawks 0.0 Heat 0.0 Team Atlantic Title Percent ==== ====================== Celtics 100.0 76ers 0.0 Team Central Title Percent ==== ====================== Cavaliers 100.0 Bulls 0.0 Pistons 0.0 Eastern Conference Playoff Advancement (Percent) Code: Team To 1st To 2nd To Conf To Fnls Champs ==== ======= ======= ======= ======= ======= Cavaliers 100.0 97.3 92.2 63.1 47.4 Celtics 100.0 94.2 61.4 24.2 14.8 Magic 100.0 91.9 37.3 11.7 6.3 Hawks 100.0 65.1 5.1 0.7 0.2 Heat 100.0 34.9 1.8 0.2 0.0 Bulls 100.0 7.6 0.7 0.0 0.0 76ers 100.0 6.3 0.7 0.0 0.0 Pistons 100.0 2.7 0.9 0.0 0.0 ======================================== Rockets Home Court Playoff Matchups By Round (Percent) Code: Team 1st 2nd Conf Final ==== ===== ===== ===== ===== Hornets 24.2 6.3 0.5 0.0 Mavericks 23.4 4.2 0.9 0.0 Blazers 12.1 0.0 4.8 0.0 Spurs 11.4 0.0 2.1 0.0 Jazz 4.8 4.4 1.0 0.0 Nuggets 0.0 15.2 0.0 0.0 Hawks 0.0 0.0 0.0 0.1 Heat 0.0 0.0 0.0 0.0 Pistons 0.0 0.0 0.0 0.0 Bulls 0.0 0.0 0.0 0.0 76ers 0.0 0.0 0.0 0.0 Rockets Away Playoff Matchups By Round (Percent) Code: Team 1st 2nd Conf Final ==== ===== ===== ===== ===== Blazers 24.0 0.0 0.8 0.0 Lakers 0.0 18.5 14.3 0.0 Nuggets 0.0 10.6 2.3 0.0 Spurs 0.0 0.0 1.5 0.0 Cavaliers 0.0 0.0 0.0 6.6 Celtics 0.0 0.0 0.0 2.5 Magic 0.0 0.0 0.0 1.3 Rockets Combined Playoff Matchups By Round (Percent) Code: Team 1st 2nd Conf Final ==== ===== ===== ===== ===== Blazers 36.2 0.0 5.6 0.0 Hornets 24.2 6.3 0.5 0.0 Mavericks 23.4 4.2 0.9 0.0 Spurs 11.4 0.0 3.7 0.0 Jazz 4.8 4.4 1.0 0.0 Nuggets 0.0 25.8 2.3 0.0 Lakers 0.0 18.5 14.3 0.0 Cavaliers 0.0 0.0 0.0 6.6 Celtics 0.0 0.0 0.0 2.5 Magic 0.0 0.0 0.0 1.3 Hawks 0.0 0.0 0.0 0.1 Heat 0.0 0.0 0.0 0.0 Pistons 0.0 0.0 0.0 0.0 76ers 0.0 0.0 0.0 0.0 Bulls 0.0 0.0 0.0 0.0 Rockets Home Court Advantage By Round Code: Round Percent ===== ======= First 76.0 Second 30.1 Conf 9.3 Finals 0.1 ================================ One game left and here are the much anticipated probabilities. Enjoy!
I understand what probabilities are Easy....but I am not all that intersted in probabilities, I am more interested in using them as predictions etc. Probabilities change as each factor changes........what is hard to quantify is how good or accurate those probabilities are in the equation. Oh I understand, they are JUST probabilities.....yeah yeah, I get it.....but then what is the point? DD
The point is you look at the western conference placement probabilities and go "oh hey, we're about even odds to get any of the 2-5 spots". You look at the odds of who we are facing in the playoffs. You look at our odds of having a homecourt matchup. People find this information interesting. To be honest, its probably not as exciting when we are 1 game away because we can pretty much figure out what the odds of certain things happening are ourselves. But back when we had like 10 games to go it was kind of cool to look and see what our odds were of winning the division according to the simulator
Agreed, but then how do you guage the accuracy of said probabilities? That is what I am trying to get at? If the answer is...you can't...then what is the point. DD
DD, this just ain't your thread. We get it, you don't see the point. There's not a lot more to be said.
i think i have to agree with DD here. if everyday...the probablility keeps changing..how is it helping us? last we had like a 49% chance of winning the southwest division...now it's like 64..that's with 1 game left..not much help if you ask me..
That's a fair criticism, maybe BigBigRed should work on a time machine so he can live these weeks 20,000 times over to see how it shakes out Really though there are ways to try and tell exactly how well the simulator has done. At first it said the chances of Utah being the 8 seed were low, but it wasn't likely at the time that they would lose almost every single road game as well as home against Golden State and Minnesota. Those were unlikely pieces of information that caused an unlikely probability to come up. Likewise the Spurs have lost some games they shouldn't have as well as sustaned some injuries. And what's the point? How many "what do you think the seeds will be?" threads have there been in the last 2 months? This one is at least consistent and a lot more fun.
things like this are not designed to be "right". it's just designed to highlight most likely possibilities. Hollingers playoff odds thing works the same way. its just a tool to guide your thinking.
What do you mean what is the point? Ask any of the people that wait for him to update the results after every game. Just because you can't know 100% the accuracy, doesn't mean you can't leave the season thinking that this simulator was solid. Look at what happened and how it happened. The two main "differences" in the first results and the last results. The Spurs were projected at #2 (which I believe they are mathmatically eliminated from getting now), and the Mavs were projected at #8. You can logically explain away those differences though. All of the other probabilities are about consistent with what happened. Which lets you draw the conclusion that this was a solid program. Whats the point you ask? Whats the point of anything? You can ask that question for about 80% of the threads on the forum. Its entertainment, its peace of mind, its interesting.
This is a pet peeve of mind, people expect methods backed by math/science/programming to be a lot more accurate than regular methods. Why? I would argue that they should be AS accurate as regular methods, but they shouldn't be expected to be more accurate.
Wait so you are looking at the accuracy of the formula too....ROFLMAO...that is exactly what I am doing. I understand the lure of the stats, I think it is cool, I am just more curious as to how the probabilities in the initial run played out. If they were highly inaccurate, then to me, it would be interesting to see what needs tweaking to up the accuracy. And yes, I understand probabilities change every day....and if all you are doing is looking at it for the probabilities, then fine...but a good many others are looking at it differently. DD
in this case there really isn't even an "accuracy" conversation to be had. when you do a simulation like this, the statements that are being made should be interpreted more like: "here are the scenarios that are possible given what we know, and of those scenarios, here's what the more likely end results look like" whether the actual result turns out to be the one that had a 30% likelihood of happening or the one that had a 2% likelihood, the simulation was still "right" and was still "accurate" because it told you it was possible in the first place. as to your comment though - you're right, but its more because stats tend to be black and white whereas qualitative datapoints are often phrased in shades of grey.
I'm not. I've already determined that the accuracy of the formula is good enough. But I'm saying that it is possible to make an educated guess as to how good the formula is using the information we know now. If the formula was crap, then the simulator would've had weird probabilities that don't really make sense. Just use your brain and determine the accuracy of the simulator yourself. When this thing started, did projecting the Spurs at #2 make sense? Even though thats not what happened, like I said, it can be logically explained why they fell like they did. But given the information we had at the time thats a logical predictions. Did projecting the Mavs at #8 make sense? Even though that might not happen, considering the bad play of Utah and the good play of the Mavs, you can see why that wasn't what happened. But again... given the information we had a month ago, it makes sense.
I understand DDs frustration. Maybe a month ago something that was said to be a 20% chance was really a 50% chance, or vice versa. There is no way to truly know how accurate it is. All you can do is use logic and make an educated guess as to how good the formula was. And DD, you keep saying that "you are curious as to how the projections play out". We are one game away from the end of the season. You can check his work right now if you want to. Does it meet your accuracy standards?
What I wonder is...How come we have a much higher chance of winning the division yet probability says we're more likely to be the 5th seed. Something about that makes not a bit of sense to me Moe
Because the requirement to win the division is that we have the tiebreaker over the Spurs and tie them or just outright beat them. Looking at the probabilities of both the Spurs and Rockets, they are mathmatically eliminated from the #2 seed and we're 30%, they have a pretty high chance to get the #5 seed due to the fact that they lose a lot of tiebreakers. We could get #2, #3 or #4 and hypothetically still have a decent shot at winning the division. Thats how we can have a decent shot at the #5 seed, yet still have such a high division winning probability.
DD is asking essential fundamental questions about how to interpret probabilities. With that in mind, comparing the actual results to the probabilities from a given date can be interesting, but interpreting them is trickier than it sounds. For example, suppose that the predictor says that we have a 4% chance of meeting the Jazz in the 1st round and then we actually meet the Jazz. That doesn't show that the predictor was wrong. It merely says that we ended up with an unlikely outcome given our assumptions about the predictor. If we want to be more quantitative about this, one could ask a question more amenable to statistical analysis, e.g. if we were to take the most likely playoff matchups predicted by the simulation, what fraction of these end up being actual matchups? Then we ask how that fraction compares to what some other prediction method would give (e.g. an example alternative predictor would be to take a survey of the matchups ppl expect here on clutchfans). Then we can compare which method does better.
Here's a quick addendum to the results from last night. The first three columns are whether or not the team wins. The fourth column is the simulator's estimated probability of the event. The last four columns are the relative placements in the Western Conference. Code: Last Game Final Rank Hou Por SA Prob Den Hou Por SA === === === ==== === === === === 0 0 0 6% 2 3 4 5 0 0 1 12% 2 4 5 3 0 1 0 11% 2 4 3 5 0 1 1 24% 2 5 4 3 1 0 0 5% 2 3 4 5 1 0 1 11% 2 3 5 4 1 1 0 10% 3 2 4 5 1 1 1 21% 3 2 4 5
The issue I have is that he says he's curious about how accurate the "predictions" are. Which implies that if they aren't acurate, then this simulator sucks (because if you are curious, and then it really doesn't matter if they are accurate or not, then you probably weren't really curious after all). And then there is the issue of measuring how accurate the simulator actually is, which as I've explained, it turns out that it really doesn't have much to do with comparing the 1st results with the last, but rather looking at the initial "prediction" and deciding for yourself based on your reasoning whether or not it makes sense. And you might say, whats the point of doing the simulator when you just test it against your own intuition anyway? The point is, once you are satisfied that the formula isn't crap, you can calculate (using math, which is never wrong, never biased), the odds of us getting a certain seed in the playoffs or us playing a certain team in the playoffs. It won't be 100% accurate, but if it passes the initial logic test, its going to be better than any non-programming system that someone else uses to determine the final seeds.
