NationStates Jolt Archive

Seeing how there are tournaments going on for Baseball, Football, American Football, Hockey, and even Russian Roulette...has anyone pondered doing a basketball tournament? Anyone interested in participating in one (if there is one?) Just wondering as it's Arroza's second sport...just behind American Football, and above Soccer and Racing.

And how the heck would you scorinate an game like that? Excel?

I think a basketball tournament would be a great idea. If would be interesting to figure out a way to scorinate the games though, im not sure how you would use excel (maybe for the standings) but im sure someone out there has some ideas on this problem. Hopefully we can become the founding fathers of a Basketball tournament in NS. Haha.

well i ran one last year but the person scorinating., ceased and they used to have a ns Cup. I would be interested. and in the last cup we had really interestig score, like 86-36 for a while then it got fixed. If we dont have enough people we could just start a chmpions league or sopmething with a few teams from each country playing international vs each other in group and knockouts.

Basketball's interesting because of the scorelines: I would imagine that a formula to generate scores would be something like a+b*rand()...A about 50 and B about 40? I don't know too much about the game and probably wouldn't participate.

Basketball's interesting because of the scorelines: I would imagine that a formula to generate scores would be something like a+b*rand()...A about 50 and B about 40? I don't know too much about the game and probably wouldn't participate.

That'd probably do for starters. The variance is slightly insane, though. And you'd probably want rand() to do more than just [0,1], say, consistently give "better" multipliers based on RP bonus.

To get fancy, you might choose to have additional madness depending on a style mod of some kind, -5 corresponding to "we own the paint" and +5 corresponding to "we live and die by the three." That's probably as complicated as it needs to get. But then you can do all sorts of crazy math, generate shooting chances and probabilities of success and *gets shot in the face for being difficult and not implementing it*

Would be a good summer project, though ^_^

Im all up for experimenting with excel to find a good Basketball scorinator. I have a ton of spare in Comp this month(School lets out in a few weeks) so I may be ale to mess around with DOS too. I sounded total nerdy, but whatever. I did some research and the average score for NBA games this year was 89.6ppg (mine as well round to 90) so maybe we can use that a base?

So 90's the base?

90* Random number between (.9 and 1.1) * Style Variance (.7 for two really defensive teams fighting it out up to 1.3 for Denver vs. Golden State) +RP Bonus (up to 10%) and Home Court (5%)?

+5: They're basically playing an And 1 open run game out there, it's all flashy offense, no D at all.
-5: Think the Bad Boys of Detroit. Slow play, rebounding, and will physically grind you into paste while winning 67-60.

Going to grind some Excel. Brb.

Sounds good so far, let us know how it works out.

First formula runout.

Random A (0) vs. Random B (0) Both without RP/Home court.

A - 106, B - 96
A - 83, B - 103
A - 92, B - 107
A - 106, B - 94
A - 108, B - 95

Random C (+5) vs. Random B (0)

C 120, B 109
C 96, B 118
C 95, B 105
C 90, B 120
C 97, B 90

Random D (-5) vs. Random A (0)

D 90, A 88
D 67, A 71
D 79, A 61
D 82, A 67
D 76, A 82

Random C (+5) vs. Random D (-5)

C 73, D 78
C 84, D 70
C 91, D 88
C 92, D 89
C 94, D 101
C 72, D 72 (Came up when I was doing other stuff!)

It's complex on how to explain the modifiers, but here's what it looks like.
A +5 team gets a modifier of 114%. That gets multiplied by the modifier of the other team. So 2 +5 teams playing against each other would get a score modifier of 130% as compared to 2 (0)'s playing. Each level would get a smaller modifier, all the way down to a -5 who's modifier would be at 84%.

Still wouldn't know where to set RP Bonuses at without driving scores up higher and higher for the really active squads. You want the active teams to win, not to drop 160 on the losing team, especially if they're defensive.

I think you could factor those bonuses into the style modifier. Rather than adding them on, you multiply them in. So it becomes mod * RP_bonus * home_court, where 'RP_bonus' takes a value from the range [1.0, 1.1] and 'home_count' takes a value that's either 1.0 or 1.05. So the 1.14 for an attacking team can turn into 1.162 if they're at home and participate fully... but at 0.84 for the defensive team also grows, to 0.97 with full participation at home. This is wrong; it should go down. And they change unevenly, too. We solve both problems by applying the multipliers to the delta.

Based on your numbers, I'm guessing that the base modifier is something like base_mod = 1.0 + delta, where delta = style mod * 0.03. We instead do delta = style mod * 0.03 * RP_bonus * home_court. That way, an attacking team with full participation sees its offense get stronger, and a defensive team with full participation will see its defense get tougher. The calculation doesn't change for the attacking team with full participation... 0.14 times all that stuff yields 0.162, which gets added in. The difference is for defense. -0.16 times all that stuff gives -0.185, so the mod goes from 0.84 to 0.815. (Still a bit uneven, but that goes away when you make +5 give 1.15 and -5 give 0.85 according to my formula.)

Given teams A and B, it seems like you have mod_A = base_mod_A * base_mod_B. Unfortunately, this means that the teams will get the full advantage of each other's RP, which is patently unfair if one team is RPing and the other is not. So team A should use its bonus-filled modifier times B's basic modifier, and B gets his bonus-filled modifier times A's basic one. We have this, and take care to note all of the primes :-)

delta = style_mod * 0.03
delta' = style_mod * 0.03 * RP_bonus * home_court
base_mod = 1.0 + delta
base_mod' = 1.0 + delta'

And you would want

mod_A = base_mod_A' * base_mod_B
mod_B = base_mod_A * base_mod_B'

I think I did all that right.

Dancougar hit the nail on the head.

What I had right now was about what Dancougar said...

Score = (Base(90) * Random between .8 and 1.2) * (Style * RP * Court.)

Actually, that base might need to be lowered slightly to 85-80 to account for the effect of there being such a high multiplier on the other side. Currently teams could average up to 135 points/game (+5 vs +5 with full rp at home)
We'll see.

I drew 8 teams, gave em all random styles, and will give them random RP points.

Testing to occur tomorrow.

Hmmm.... well, not quite. If defensive team A roleplays a lot, and offensive team B doesn't roleplay at all... offensive team B doesn't feel the full power of A's defense, and it should. I seem to have introduced another imbalance.

Blaugh. This is why I prefer simulation to a straight formula ^_^ But that's even more work, and kind of misses the point of an RP tournament.

I was crunching some numbers running a default style modifer (ie. A flat score that adds to your score or subtracts from opponent). The equation I used to find the score was (85(base)*(.5-1.2))*(RP Bonus (.1-1.5)+(.5 or 1 being Homecourt) heres a sample equation.

Team A (Playing an Offensive +4 team)
(85*.7)*(.6+1) (1 being that team has home court advantage)
59.5*1.6
95.2 Un-modified score

Style Bonus for Team A
(+4*2) = +8

Team B (Playing a Defensive -3 team)
(85*1.1)*(.7+.5)
93.5*1.2
112.2

Style Bonus for Team B
(-3*2) = -6

Final scores being
TEAM A
95.2 + 8 - 6 = 97
TEAM B
112.2 + 0 (No Modifiers affecting this team) = 112

Thus Team A wins 112-97 A realistic score in today's NBA.

If we can find a way to influence the (.8-1.2) depending on the persons RP Bonus and Homecourt we could balance the scores much easier. This method requires more work but in my opinion the style modifiers don't play to much of a role.

Hmmm.... well, not quite. If defensive team A roleplays a lot, and offensive team B doesn't roleplay at all... offensive team B doesn't feel the full power of A's defense, and it should. I seem to have introduced another imbalance.

Blaugh. This is why I prefer simulation to a straight formula ^_^ But that's even more work, and kind of misses the point of an RP tournament.

So RP'ing a strong Defense, would cause the scores to go down, not up. And preferably in some way that rewards the RP'ing squad.

I'm going to be pondering that all the way to Montgomery tomorrow. I had that formula anyways, so don't feel bad about introducing an imbalance. But there should be someway to introduce negative modifiers, while still giving the better team an advantage. Maybe reciprocals....

So RP'ing a strong Defense, would cause the scores to go down, not up. And preferably in some way that rewards the RP'ing squad.

I'm going to be pondering that all the way to Montgomery tomorrow. I had that formula anyways, so don't feel bad about introducing an imbalance. But there should be someway to introduce negative modifiers, while still giving the better team an advantage. Maybe reciprocals....

I hate reciprocals, but I'll have a chat with some of the CompSci nerds today maybe they can come up with something.

I was crunching some numbers running a default style modifer (ie. A flat score that adds to your score or subtracts from opponent). The equation I used to find the score was (85(base)*(.5-1.2))*(RP Bonus (.1-1.5)+(.5 or 1 being Homecourt) heres a sample equation.

Team A (Playing an Offensive +4 team)
(85*.7)*(.6+1) (1 being that team has home court advantage)
59.5*1.6
95.2 Un-modified score

Style Bonus for Team A
(+4*2) = +8

Team B (Playing a Defensive -3 team)
(85*1.1)*(.7+.5)
93.5*1.2
112.2

Style Bonus for Team B
(-3*2) = -6

Final scores being
TEAM A
95.2 + 8 - 6 = 97
TEAM B
112.2 + 0 (No Modifiers affecting this team) = 112

Thus Team A wins 112-97 A realistic score in today's NBA.

If we can find a way to influence the (.8-1.2) depending on the persons RP Bonus and Homecourt we could balance the scores much easier. This method requires more work but in my opinion the style modifiers don't play to much of a role.

In this case the style modifiers are imbalanced though. What team wouldn't take a +5 over a -5 when it gives you an extra 20 ppg, and the modifier doesn't affect the opponent. When a team plays breakneck offensively, it should up the score for both teams, as both teams get many more possessions.

Team Score = ( Base (90) * Randomizer ) * (Game Modifier * RP Bonus * Court Advantage.)

The question then is, how to modify the game to suit the style of the team that rp's more.

Game Modifier = ( (Team A Style * RP Bonus) * (Team B Style * RP Bonus) )

Well that works for offensive and neutral teams. When they RP they get a bonus to the modifier that makes the game more offensive. What if Team B's defensive?

Game Modifier = ( (Team A Style * RP Bonus) * (Team B Style * (1-RP Bonus)) )

That would make more defensive teams that RP have more defensive games overall, right?

Now if only all this could be simplified.

In this case the style modifiers are imbalanced though. What team wouldn't take a +5 over a -5 when it gives you an extra 20 ppg, and the modifier doesn't affect the opponent. When a team plays breakneck offensively, it should up the score for both teams, as both teams get many more possessions.

Good point there, as I said this is just starting stages. Maybe making my style bonus to give to both teams (+2 would give both teams 4 points, a -3 would take 6 points off of both teams scores).

Now that I think about thats probably even worse.

Reads Dancougar's post.

Reads Dancougar's post.

Reads Dancougar's post.

Damn...you're a genius.

Dancougars onto something here, his idea actually does work after crunching some numbers. Though I don't get why one team should get another teams RP bonus, especially if one of those teams never even posted a roster.

If teams don't post roster (or atleast a style modifier that team is playing) should it be an automatic zero to the RP and style modifiers?

The other thing is everyone now and then you see upsets liek BIG upsets how can do these further on down the road once we have some established teams (further taking into account experience). I mean no one saw the Hawks taking the Celtics to game 7.

Anyway lets just focus on this first tournament then figure out ideas for future use (like a league simulation using RPing).

Reads Dancougar's post.

Reads Dancougar's post.

Reads Dancougar's post.

Damn...you're a genius.

As I said, my formula still doesn't give the defensive mod the full benefit of the RP bonus.

The case that bothers me is when the offensive team doesn't RP and the defensive team does. The calculation for the defensive team's mod uses the bonus-ful style mod as it should, but the offensive team doesn't... and so from his point of view, the defense's delta is a smaller negative number. He faces a weaker defense than he should.

The case where the offense RPs and the defense doesn't works as expected.

I think that's the only bad case, and all you would have to do is do the calculation with bonus-ful style mod and not, and take the smaller result. That's still kind of messy and I don't like it because it makes assumptions about the relative value of each strategy.

Dancougars onto something here, his idea actually does work after crunching some numbers. Though I don't get why one team should get another teams RP bonus, especially if one of those teams never even posted a roster.

They don't. Team A mod = Team A base mod with bonuses * Team B base mod without bonuses

If teams don't post roster (or atleast a style modifier that team is playing) should it be an automatic zero to the RP and style modifiers?

Right. Which means they get the default 1.0 for a 0 style mod, and the default 1.0 for no RP. If we allow "negative" RP, then the RP bonus multiplier would dip below 1.0 and shrink their bonus. But to do that, in addition to figuring out a rubric for positive RP score, you need to define levels of penalization, which is kind of nastier.

One thing. If I'm a defensive team, it's actually in my interest not to rp.

When you do, the rp bonus gets multiplied by the negative modifier, making you play more negatively.

I think I did something to it to make it work, by having both teams have the same modifier (which reflects the overall pace of the game). Then having the RP bonus, and home court be added to each team again seperately.

That way if a RP'ing defensive team plays a non rping offensive team, the score should be lower than if neither had rp'ed. And vice versa. Anyways it's 0100, and I'm at a truck stop in Knoxville...time to pass out.

It should work. I'll explain tomorrow.

The problem as I see it is that style mods and RP-bonus seem too intertwined from what's being discussed so far. Apologies for stuff people probably brought up already:

RP-bonus: can't it simply be multiplied in for each team independently? Two teams, A and B, A RPs, B doesn't-I think it's OK for A to score more than B, even if both teams are defensive. RP-bonus is supposed to make one do better, but that doesn't need to correlate with someone's style. A team that scores more points will do better than one that doesn't?

As far as style modifiers go, should anything more than their sum* influence the overall game? An offensive team and an equally defensive team should have the same scoreline as two neutral teams? I don't think which team RPs more needs to have anything to do with it.

*Or product, I guess, if you had a range from like .7 to 1.3-but the extremes would need to multiply to 1, not add to 2, I think.

The problem as I see it is that style mods and RP-bonus seem too intertwined from what's being discussed so far. Apologies for stuff people probably brought up already:

RP-bonus: can't it simply be multiplied in for each team independently? Two teams, A and B, A RPs, B doesn't-I think it's OK for A to score more than B, even if both teams are defensive. RP-bonus is supposed to make one do better, but that doesn't need to correlate with someone's style. A team that scores more points will do better than one that doesn't?

As far as style modifiers go, should anything more than their sum* influence the overall game? An offensive team and an equally defensive team should have the same scoreline as two neutral teams? I don't think which team RPs more needs to have anything to do with it.

*Or product, I guess, if you had a range from like .7 to 1.3-but the extremes would need to multiply to 1, not add to 2, I think.

What Zuangzug said.
Simplicity = bliss, even if it means defensive teams score slightly more.

Good points all around. I think my fundamental misunderstanding is my interpretation of what it means to play a defensive strategy. It should hurt the other team's ability to score, but not your own. And an offensive strategy should aid your ability to score, not necessarily make it harder for opponents to stop you.

Yeah, I'd call something that decreases the other team's ability to score but not your own a bonus of sorts, rather than a style modifier. Style modifiers reflect the overall strengths of a team, but by extension the weaknesses too. A defensive team should concede less, but also score less, than an offensive one of equal caliber.

And an offensive strategy should aid your ability to score, not necessarily make it harder for opponents to stop you.The first half is true, but I'm not sure how you'd code "making it harder for opponents to stop you" any differently. But an offensive style should also allow your opponents to score more-I guess it's a difference between strategy and style modifiers the way they're traditionally understood. Unless I've missed the point too, which is quite possible :p

Yeah, I'd call something that decreases the other team's ability to score but not your own a bonus of sorts, rather than a style modifier. Style modifiers reflect the overall strengths of a team, but by extension the weaknesses too. A defensive team should concede less, but also score less, than an offensive one of equal caliber.

The first half is true, but I'm not sure how you'd code "making it harder for opponents to stop you" any differently. But an offensive style should also allow your opponents to score more-I guess it's a difference between strategy and style modifiers the way they're traditionally understood. Unless I've missed the point too, which is quite possible :p

I think what we've both figured out is that finding a single formula to cover the expected behavior for defensive style mods and offensive style mods is very tricky :-)

I'm not interested, but...

If you want realistic scoring based off NBA scores, it's not enough to know the average score. You'll also need to know the standard deviation--let's assume that NBA scores follow a normal distribution.

I'm not interested, but...

If you want realistic scoring based off NBA scores, it's not enough to know the average score. You'll also need to know the standard deviation--let's assume that NBA scores follow a normal distribution.

Forget I ever asked.

I'm not interested, but...

If you want realistic scoring based off NBA scores, it's not enough to know the average score. You'll also need to know the standard deviation--let's assume that NBA scores follow a normal distribution.

Hmm, maybe that's a better way to apply the RP and home court bonus. Instead of altering the style modifier, alter the base score. Say a good RP bumps you up a sigma or two. Not even that, we can just pick a flat number. Contrived example, a team with max RPs starts off at maybe 100 * style_mod instead of 90. That way, RPing ALWAYS increases your chance of scoring.

And the nice thing is that you can just use the original formulas you had for style modifiers, none of my delta nonsense :-D

I would be interested in participating.