-
Welcome to Smogon! Take a moment to read the Introduction to Smogon for a run-down on everything Smogon, and make sure you take some time to read the global rules.
-
Congrats to the winners of the 2023 Smog Awards!
LCPL 5 Discussion
- Thread starter Camden
- Start date
pls draft me to save me from dota
Yours Truly,
PO Tier Leader Fitzy
(i actually don't want to get involved believe it or not I'm indifferent where the auction is)
I actually love using IRC and i'm on it whenever im at my computer, but that isn't the case for the majority of the community anymore. having the draft on PS makes it more accessible for newer users that really don't know how IRC works or even what IRC is. If there is a problem with lag on the day of the draft, OP still has the auction bot fully functional on IRC and I still have #lcplauction registered so it would take maybe 5 minutes to transfer the draft to there on the day of the draft, but PS just seems to be a lot more accesible for a majority of the community and thats why I decided to have the draft there.
after the managers and i had an enlightening conversation im being told to make dpp a bo1 again, sorry to whoever i mislead and hopefully nobody is too mad over my decisions and if you are, feel free to pm me if you want to yell at someone
Last edited by a moderator:
can they post a valid, well-reasoned, explanation actually addressing the issues left unanswered in the thread?
....or does that take too much time to prep as well?
logically:
1) we play this game to win (no shit)
2) for a game to be competetive (and thus have who wins [see: 1)] matter) the better player must win
3) In pokemon, like most games, there is a luck-based element that might lead to a "worse" player winning a given game
4) multiple games in a set increase the likelyhood of the better player prevailing by providing a large sample to mitigate the potential effects of luck
4a) this means better players win more often and the outcome is more meaningful based on who wins being important
5) thus larger samples are preferable in pokemon games.
the reason NOT to have larger samples is it is hard - to build, to find time, etc. u have prolly the best dpp player and a seemingly busy dude saying it doesnt take a ton of time to build practice and play dpp, meaning it's preferable to have more games, as it always is.
1) we play this game to win (no shit)
2) for a game to be competetive (and thus have who wins [see: 1)] matter) the better player must win
3) In pokemon, like most games, there is a luck-based element that might lead to a "worse" player winning a given game
4) multiple games in a set increase the likelyhood of the better player prevailing by providing a large sample to mitigate the potential effects of luck
4a) this means better players win more often and the outcome is more meaningful based on who wins being important
5) thus larger samples are preferable in pokemon games.
the reason NOT to have larger samples is it is hard - to build, to find time, etc. u have prolly the best dpp player and a seemingly busy dude saying it doesnt take a ton of time to build practice and play dpp, meaning it's preferable to have more games, as it always is.
Ya I don't see a reason why there can't be Bo3 in DPP. It honestly does not take long to build a decent team at all. I am totally opposed to this being applied to any other tier, but DPP is just so fast-paced and not hard to build for. It's disappointing to only play one match with so little turns.
Stop putting bo3 in team tourneys. If you're going to put bo3 in a tier, you have to put in all of them. Otherwise you're having just one tier that "the best player will win" more often and that's totally bullshit. All the tiers should be equals Oo
Also, time to build/duration of a match is only relevant if you can't do both in a week.
Also, time to build/duration of a match is only relevant if you can't do both in a week.
HANTSUKI Heysup rozes Shrug
this is the discussion thread
Now, that being said, I'm not vehemently supporting a Bo3 or a Bo1, as, in truth, it is unlikely that I (or, to be fair, most of us) will be playing DPP. I just know that, as a DPP spectator, I would love to see the tier played out to an equal extent as other tiers
this is the discussion thread
They made RBY best of three in SPL in order to mitigate extraneous factors (in this case, severe hax potential) that made doing one game a poor decision. DPP LC is not nearly so hax-riddled; my main reason to push for a Bo3 is the length of the games. Several times last LCPL, I would go to play an LC UU match with plans to watch my teammate Raseri play after I lost. Even when Raseri started playing long after I did he would finish before I ever had the chance to watch. DPP LC is a heavily offensive tier based largely on revenge killing, making good DPP LC games incredibly short, and therefore pivotal on one or two turns. I'd like to be able to watch more DPP, and making DPP a Bo3 will likely extend each series length to somewhere near the length of one good BW2 match; this would also help make the matches across each generation equal measures of Pokemon skill.
Now, that being said, I'm not vehemently supporting a Bo3 or a Bo1, as, in truth, it is unlikely that I (or, to be fair, most of us) will be playing DPP. I just know that, as a DPP spectator, I would love to see the tier played out to an equal extent as other tiers
I don't think "I wanna watch more matchas because they're so fast!!!!" is something you should think about while deciding if a tier should be bo3 or not Oo
And even if you think about this, you should let other spectators watch more games of the other tiers they like as well because they could say "hey, I wanna watch some good series where ~the best player really wins~" instead of a bo1 where a crit/roll/speed tie/para/scald burn just decides the game. You can say you wanna more time of ur tier, but they can say they want quality. So everything should be bo3 or bo1 :(
Also the RBY thing is just bullshit as well, because you need A LOT OF HAX to change a game in RBY while some 30% shit just fucks a lot of games in any new gen...the hax argument should make all tiers bo3 if applied, but not talking about this here :O
And even if you think about this, you should let other spectators watch more games of the other tiers they like as well because they could say "hey, I wanna watch some good series where ~the best player really wins~" instead of a bo1 where a crit/roll/speed tie/para/scald burn just decides the game. You can say you wanna more time of ur tier, but they can say they want quality. So everything should be bo3 or bo1 :(
Also the RBY thing is just bullshit as well, because you need A LOT OF HAX to change a game in RBY while some 30% shit just fucks a lot of games in any new gen...the hax argument should make all tiers bo3 if applied, but not talking about this here :O
While, like I said, I would love to see more DPP be played, that's not actually my main point. I guess I really didn't make it clear, but making DPP a Bo3 represents the tier equally via actual turns played and play time with the other two tiers. DPP is without a doubt the fastest meta represented in LCPL, and this is largely due to the absence of bulk-boosting items such as Eviolite and Berry Juice. The main reason this should come into play in LCPL is the nature of level 5 stats. The length of SPL games is relatively similar across generations bar GSC most of the time (but that one is actually longer so it doesn't apply to the argument); this is due to level 100, fully evolved stats being far more balanced and bulky than their level 5 counterparts. In generation 5, Eviolite was introduced, which for the most part mitigated the disproportionate bulk issue from there on out. Because of this, DPP games are much shorter than later gen matches.
I suppose my posts here are trying to propose a few questions to consider that can settle the issue, and I think it boils down to these:
1) Is the difference in length of DPP games significant in determining players' skill?
2) Is an average DPP match of about 14-15 turns (from previous replays averages) as representative of players' skill as 26 turn BW2 matches and/or 21 turn ORAS matches?
3) Is the community's demand for more DPP so incredibly large as to warrant more matches?
If the answer to 1 is no, then 2 is irrelevant and it's likely that Bo1 would be the best option. If 3 is yes, which would require massive community support, the Bo3 is the best option. If we determine that 1 is yes, then we turn to 2 to determine whether that difference is significant enough to warrant extending the series to Bo3.
So we have actual numbers to work with, I'll do the number of turns for DPP games last lcpl.
Week 1: 20, 11 (due to inactivity from Silverghost, both sides had 5 or more Pokemon left meaning a minimum of a 16 turn game), 18, unknown
Week 2: 15, 22, 17, unknown
Week 3: 27, 24, 26, 20
Week 4: 19, 15, 16, 14 (forfeit before loss, would have ended on turn 15)
Week 5: 13, 34, 20, 20
Week 6: 11, 18, 15, unknown
Week 7: 14, 25, 19 (forfeit before loss, would have ended on turn 20), unknown
Playoffs: 19, 22
So the shortest game lasted 11 turns while the longest lasted 34. The average number of turns (unaltered, so forfeits and the Silverghost battle are as written) is 19. The minimum actual average (assuming all forfeits play out to last turn and SG battle somehow ends on turn 16) is 19.269 turns.
I don't want to take a stance on this because I don't play DPP (unless required of me by my manager) but I figured I'd take a little time to give some actual numbers to what's being stated. I might do ORAS as well, to give a comparison point.
Also please note that number of turns shouldn't be considered a measure of skill.
Week 1: 20, 11 (due to inactivity from Silverghost, both sides had 5 or more Pokemon left meaning a minimum of a 16 turn game), 18, unknown
Week 2: 15, 22, 17, unknown
Week 3: 27, 24, 26, 20
Week 4: 19, 15, 16, 14 (forfeit before loss, would have ended on turn 15)
Week 5: 13, 34, 20, 20
Week 6: 11, 18, 15, unknown
Week 7: 14, 25, 19 (forfeit before loss, would have ended on turn 20), unknown
Playoffs: 19, 22
So the shortest game lasted 11 turns while the longest lasted 34. The average number of turns (unaltered, so forfeits and the Silverghost battle are as written) is 19. The minimum actual average (assuming all forfeits play out to last turn and SG battle somehow ends on turn 16) is 19.269 turns.
I don't want to take a stance on this because I don't play DPP (unless required of me by my manager) but I figured I'd take a little time to give some actual numbers to what's being stated. I might do ORAS as well, to give a comparison point.
Also please note that number of turns shouldn't be considered a measure of skill.
I understand that people think Bo3 in team tours should be discouraged because of the nature of it potentially bogging down the tournament or removing the excitement (which I think is 73.4% arbitrary subjective bullshit). And even though I disagree with it, RBY in SPL has already set the precedent for allowing Bo3s in one tournament over other areas. I would argue all of the cons from making RBY bo3 completely fade away in DPP as well since it's not going to be a 3 hour haxfest it's going to be 15 minutes of fast-paced games. While this isn't about "skill" or "making sure the best player really wins", I would argue they are a bonus or secondary benefit of making DP bo3.
I would make the analogy to Super Smash Brothers:
Oras and BW are 1v1 stock 1 game where as DPP is a sudden death stock 1 game (essentially if you're hit once you die). Sure, the 1 stock regular game is only one life, but that one life takes a reasonably wide array of attacks before being at a high enough percent to be knocked out. In the Sudden Death game, it still takes skill to dodge and avoid that one hit, but it is anti-climactic and almost meaningless to win based on one hit. What rational person would suggest that a tournament holds any weight if it is based on just a 30 at best second smash game when the rest of the games are easily 2 minutes at least?
I have already addressed the prep time issue along with the relevant people who commonly play and build for the tier. I have also addressed why DPP is different from the other tiers, so it would be asinine to still claim that there's no good reason why DPP deserves a bo3 and other tiers don't. In case you forgot/didn't read: it requires less prep time and and the games take less time to the point where the sum of 2-3 games will likely paint a more similar picture of the level of prep and games as in ORAS/BW.
And I don't want to hear "if it ain't broke don't fix it". People noticed this problem when they played the forum DPP tournament. It seems to me that if the rule was fixed there, the issue is clearly relevant. Read this thread. I mean if you think this conversation and this many people agreeing would be happening without an issue then \_0_/?
Seriously, Diglett was fucking suspected with less of a public request than this (ya, I'm exaggerating here but not by as much as we would all like to admit).
brb trying to style star.
I would make the analogy to Super Smash Brothers:
Oras and BW are 1v1 stock 1 game where as DPP is a sudden death stock 1 game (essentially if you're hit once you die). Sure, the 1 stock regular game is only one life, but that one life takes a reasonably wide array of attacks before being at a high enough percent to be knocked out. In the Sudden Death game, it still takes skill to dodge and avoid that one hit, but it is anti-climactic and almost meaningless to win based on one hit. What rational person would suggest that a tournament holds any weight if it is based on just a 30 at best second smash game when the rest of the games are easily 2 minutes at least?
I have already addressed the prep time issue along with the relevant people who commonly play and build for the tier. I have also addressed why DPP is different from the other tiers, so it would be asinine to still claim that there's no good reason why DPP deserves a bo3 and other tiers don't. In case you forgot/didn't read: it requires less prep time and and the games take less time to the point where the sum of 2-3 games will likely paint a more similar picture of the level of prep and games as in ORAS/BW.
And I don't want to hear "if it ain't broke don't fix it". People noticed this problem when they played the forum DPP tournament. It seems to me that if the rule was fixed there, the issue is clearly relevant. Read this thread. I mean if you think this conversation and this many people agreeing would be happening without an issue then \_0_/?
Seriously, Diglett was fucking suspected with less of a public request than this (ya, I'm exaggerating here but not by as much as we would all like to admit).
brb trying to style star.
Last edited:
-_-
I'm gonna clarify this
Since Merritt did the DPP and my numbers were off, I redid my BW ones.
Week 1: 34 33 25 35
Week 2: 27 37 31 34
Week 3: 51 22 37 30
Week 4: 24 25 34 24
Week 5: 40 12 30 X (SLAM CLAM)
Week 6: 28 20 25 24
Week 7: 40 30 X X
Semis: 25 26
Finals: X. (not in the thread)
The average length of these games came out to be 29.47 turns, about 1.5 times as many as the average DPP game. Making the DPP series a best of 3 would essentially double the number of turns played in optimal scenarios, so that'd be around 38ish turns per series. It would seem that that is way more DPP turnwise, but keep in mind that 35-40 turn BW2 games are not rare occurrences by any means.
What I want to clarify is the "turns played doesn't equal skill thing" because, even though that's not what I really meant, I'm pretty bad at explaining my thought process I think. What I was trying to say in terms of the questions I asked is that those extra turns, which are a result of the overall playstyle and metagame of BW2 and not of player skill, are an extra window in which better players can demonstrate their skill. BW2 players play more Pokemon than DPP players in almost every single case, sometimes by substantial margins. My question is whether or not those extra turns of play are a significant enough difference to warrant extending the DPP series. If not, cool! I really don't mind either way. Would I like to see more DPP cause it's fun to watch? Of course! Would it kill me if we kept the Bo1? Of course not. Good LCPL DPP players such as Heysup and Kumiho and other community members such as Quote and Shrug have given reasons supporting a Bo3, and most people's objections have been addressed (mostly by Heysup) so what would keep this from happening?
I find this a rather useless subjective type of argument/discussion, despite the good points made and several different evidence for both sides. But what I do feel needs to be stated is that technically from a lineup standard oras is a bo3 each week, because 3 oras starters, while the rest are bo1. Whether or not you like it scouting and whatever the hell Mambo does (stalking?) Is also a sign of being the better play, hes being the smarter player if his team works.
Thank you Berk for the BW, I had been halfway done with them ;_;
ORAS turn numbers.
Week 1: 16, 38, 22, 52, 29, 36, 28, 35, 33, 58
Week 2: 23, 38, 8 (Levi brought BP, minimum number of turns before end was 15), 27, 26, 47, 32, 22, 25
Week 3: 27, 25, 28, 25, 26, 30, 25, 24, 27, 30, 28, 35
Week 4: 21, 16, 24, 22, 17, 29, 35, 29, 25, 33
Week 5: 19, 26, 23, 33, 27, 23 (forfeit, would end 24), 27, 36, 22, 24, 19
Week 6: 21, 26, 21, 18, 13, 36, 42, 22, 29
Week 7: 28, 18, 23, 11, 33, 11 (forfeited, would end 14), 20, 33, 32
Playoffs: 18, 22, 22, 23, 30, 21
Finals: 23, 37, 21
The shortest game (ignoring KM's forfeit against Levi) was 11 turns while the longest was 58, meaning that ORAS tied with DPP for shortest game and beat out BW for longest. The average number of turns unaltered was 26.696 turns. The minimum real average number of turns was 26.835. The median number of turns was 26. The fact that the mean and median are almost identical is quite interesting.
Using Berk's analysis of BW, when we compare the average (unaltered) of all three, it results in BW being the longest on average (1.55x longer than DPP and 1.10x longer than ORAS), ORAS being the middle child (1.41x longer than DPP and .91x as long as BW), and DPP being the shortest by far (.64x as long as BW and .71x as long as ORAS). Beyond the argument, this is also interesting in reflecting what playstyle dominates each meta, where BW has a much defensive metagame than DPP's hyper offense and ORAS's more balanced.
It's also worth noting that all three have their own outliers from the average, DPP having a 34 turn game, BW having a 12 turn game, and ORAS having both an 11 turn game and a 58 turn game.
ORAS turn numbers.
Week 1: 16, 38, 22, 52, 29, 36, 28, 35, 33, 58
Week 2: 23, 38, 8 (Levi brought BP, minimum number of turns before end was 15), 27, 26, 47, 32, 22, 25
Week 3: 27, 25, 28, 25, 26, 30, 25, 24, 27, 30, 28, 35
Week 4: 21, 16, 24, 22, 17, 29, 35, 29, 25, 33
Week 5: 19, 26, 23, 33, 27, 23 (forfeit, would end 24), 27, 36, 22, 24, 19
Week 6: 21, 26, 21, 18, 13, 36, 42, 22, 29
Week 7: 28, 18, 23, 11, 33, 11 (forfeited, would end 14), 20, 33, 32
Playoffs: 18, 22, 22, 23, 30, 21
Finals: 23, 37, 21
The shortest game (ignoring KM's forfeit against Levi) was 11 turns while the longest was 58, meaning that ORAS tied with DPP for shortest game and beat out BW for longest. The average number of turns unaltered was 26.696 turns. The minimum real average number of turns was 26.835. The median number of turns was 26. The fact that the mean and median are almost identical is quite interesting.
Using Berk's analysis of BW, when we compare the average (unaltered) of all three, it results in BW being the longest on average (1.55x longer than DPP and 1.10x longer than ORAS), ORAS being the middle child (1.41x longer than DPP and .91x as long as BW), and DPP being the shortest by far (.64x as long as BW and .71x as long as ORAS). Beyond the argument, this is also interesting in reflecting what playstyle dominates each meta, where BW has a much defensive metagame than DPP's hyper offense and ORAS's more balanced.
It's also worth noting that all three have their own outliers from the average, DPP having a 34 turn game, BW having a 12 turn game, and ORAS having both an 11 turn game and a 58 turn game.
After reading this whole treath and seeing the numbers, I wanted to analyse the data from Merritt and H&MBerkerly with statistics learned by Analytical Chemistry and Bioanalysis (chemistry batchelor second year). I have first calculated the mean/average and the standard deviation using excel (I tried to add it in the appendix but I can't upload spreadsheets so if you want it, so can contact me and I will email it!). Then using a Grubbs-test I have tested different outliers and discarted when they changed the standard deviation too much. This means the values in the ORAS data of 52 and 58 turns are now discarted as well as the 34 turn game of DPP. Then I calculated the new average and new SD but now I also calculated the Induced-SD (this uses the fact that more data means higher probability of having the right value). Then I used the Student-T value to do get the 95% confidence interval of the amount of turns. This confidence interval says that there is a 95% chance of finding the amount of games in this interval. This gives for ORAS, 26.10+/-1.57 turns, BW2: 29.40+/-3.03 turns and DPP: 18,64+/-1,66. This means, rounded down to natural numbers, that the number of games played in ORAS is between 25 to 28, BW2 26 to 32 and DPP 17 to 20. This means that DPP games are shorter (what a surprise!!!) and ORAS and BW2 are almost as long.
I hope that this can add something to the discussion because these are, when I didn't make a mistake, the true values and nobody can deny those numbers. Maybe the more experience LC players can use these data to defend their points because I will not give my verdict. This may also be redundant and I may come over as a math nerd but I wanted to give something to the LC community with something what I'm good at and like. So don't hate and don't laugh!!!
I'm also objective because I didn't sign up for the LC DPP metagame and only for the ORAS one for LCPL. The only things I have behind my name is that I had suspect rights for the Diglett-retest in 55 games with 78% GXE and I laddered to the top 100 last weekend (1515) GXE: 83 so this is also a reason why I will not give my opinion.
So again I hope this is a good contribution and you will enjoy my work!
I hope that this can add something to the discussion because these are, when I didn't make a mistake, the true values and nobody can deny those numbers. Maybe the more experience LC players can use these data to defend their points because I will not give my verdict. This may also be redundant and I may come over as a math nerd but I wanted to give something to the LC community with something what I'm good at and like. So don't hate and don't laugh!!!
I'm also objective because I didn't sign up for the LC DPP metagame and only for the ORAS one for LCPL. The only things I have behind my name is that I had suspect rights for the Diglett-retest in 55 games with 78% GXE and I laddered to the top 100 last weekend (1515) GXE: 83 so this is also a reason why I will not give my opinion.
So again I hope this is a good contribution and you will enjoy my work!
The main question I have about the numbers is whether or not you used the exact turn values I gave or the minimum actual turn values? This means taking into account the forfeits and disconnects that made the turn number lower than expected. BW only had one of these, the SLAM CLAM game, where it ended on turn 12 but would have actually finished on turn 15 due to Eternal Spirit having 3 OHKOed mons left. DPP and ORAS had more of these overall (3 in both DPP and ORAS) which should both boost their average up a bit, and actually helps out with the standard deviation since it means you no longer have the absurdly low minimum of 8 in ORAS and DPP only has one game at 11 turns rather than 2.
Overall though I was hesitant to do this kind of analysis due to the nature of different playstyles. HO vs HO results in drastically shorter games than Stall vs Stall, which is a large part of why there's so much variance and the longest games are so many times longer than the shortest ones. The most accurate way I could see to portray it fairly would be actually separate the games into groups based on what playstyle each team was using, but there's a subjective aspect of this that makes it difficult (also that would require looking at movesets and teams rather than just the last turn so it'd take much more time). Overall though I do like what you did with the numbers, and the values you got are interesting!
