EPT
. Ace Goojila I believe and ace great !Fantasy Proleague 11-12 R1 Signup/Discussion - Page 13
| Forum Index > BW General |
|
Sawamura
Malaysia7602 Posts
. Ace Goojila I believe and ace great ! | ||
|
Phelix
1931 Posts
Main Team 4  BeSt1  Canata1 hOn_sin6 Light7 Sea8  ZerO (captain)3 Woongjin Stars Anti Team 5 Action2 Last6 SnowMy prediction is that Woongjin will cause upsets this year due to the format, and thus I am heavily investing in them with Light, Zero, and hOn_sin (Woongjin coach will send him out a lot more, and the maps do favor Terran according to the news article). Canata because he's my favorite player, and ACE's choices for lineups amazes me sometimes. Best for SKT T1 win points, as well as lineup chances as they always seem to dominate Proleague, especially Fantasy Proleague Formats. Sea for Team 8, as well as always being a solid contributor in Proleague. Action will not be sent out much if he keeps slumping like this, as KT is just way too deep for slumping players. Snow from CJ because he's on a team with too much depth. Both of them will cause team win points, which will be a slight hurt to my team. Last from STX because his team is going to be dead last, and thus will not score points. One thing I regret is that bad there's no Lomo to put on my Anti-Team anymore. | ||
|
Xiphos
Canada7507 Posts
Main Team 6 BaBy5 EffOrt (captain)4 firebathero6 Neo.G_Soulkey2 Paralyze4 Shuttle3 Samsung KHAN Anti Team 5 Action4 Crazy-Hydra4 s2I am VERY keen on Effort's revival. | ||
|
Tufas
Austria2259 Posts
Should I go 30/30 with great missing (with him I would have all 5 main ACE players) or take 28/30 with great (all 5 together are 18 points) Should I try to make a lot of points (well I have 4 ACE players anyway ..) or go for the all in with great Poll: Should I let great into my main team even if that means 28/30 points ? YES, as a ACE fanboy you have to do this. (13) No, 2 points wasted are really not in the spirit of the FPL. (6) 19 total votes Your vote: Should I let great into my main team even if that means 28/30 points ? (Vote): YES, as a ACE fanboy you have to do this. | ||
|
Ikonn
Netherlands1958 Posts
| ||
|
figq
12519 Posts
As promised, some details: + Show Spoiler + Turns out generally the more expensive players are less cost-efficient. At the bottom of the list are Jangbi, Jaedong, Fantasy, Stork, and at the top are Alone, BByong, yeOngJae and other 1-point players from good teams. For team I use Team 8, because it has the strongest top 5 lineup. As estimate for its wins I use 27, compared to 28-30 for CJ, KT, SKT, which actually seems a bit conservative. For antiteam, I originally wanted to center it around Stork, but the results convinced me that Jangbi combo is the better option, just because he doesn't play as much, and he doesn't win even half the matches of Stork's. Moreover, I won't be surprised if Jangbi goes into complacency mode now and not push himself too much in the Proleague. For the team, the only major change based on the calculation was to pick less distributed team representation, namely 3 SKT and 2 Stars players. Originally I was planning a very similar lineup, but with Effort from CJ and a couple other representatives from different teams. But eventually the stats convinced me. Because as long as SKT, KT and Stars win stable amount of wins, they just bring more points than a distributed configuration. I guess I'll have to hope that CJ fails. Its good players are just a little too expensive, and the bad players aren't worth it. For Team 8, at least based on last Proleague, none of their players were cost-efficient enough to cut it. On the other hand, I will have the team itself as my main team, so even if they go on a rampant stomping, I won't be completely screwed, and the other big teams will be able to score good results in the rest of the matches. So the team: http://www.teamliquid.net/fantasy/proleague/Team.php?team=figq + Show Spoiler + Main Team 1 Alone (SKT) 7-13-- Strong because he's on SKT, and he actually gets to play a lot, for a 1-pointer. 4 BeSt (SKT) 19-12-- most cost-efficient 4-pointer, based on last PL (he was on the lineup two more times than Soo) 4 Crazy-Hydra (KT) 17-20-- good enough, and from KT; note that even if he loses quite some games, he just racks a high number of wins, which matters 6 Light (WJN) 28-16-- plays a ton and wins a lot, worth his 6 points 4 n.Die_soO (captain) (SKT) 19-10-- very solid, I don't expect to be changing him, so he's captain 6 Neo.G_Soulkey (WJN) 28-15-- maybe 2 Stars players is a stretch, but both of them just deliver too much FPL points to be excluded, and so the optimization procedure picks them over others 5 Team 8 - they have slightly better top 5 than the rest, even though much shorter bench. Anti Team 6 Jangbi (SAM) 13-18-- I hate to put him on antiteam, but the facts are facts, he's the most cost-inefficient player based on his results from PL 4 Kal (ACE) 14-21-- not bad, but most cost-inefficient 4-point player, even though I expect him to get better, just suits the space too well 3 sHy (WJN) 3-4-- a bit contradictory, because of the team, but his lack of performance compared to cost just compensates for it. That's where the algorithmic approach might get screwed, because as a young player he could get a lot more used this season than the previous one, but hey, I want to see how this little test will end up. And here's a full list of players: Cost / Race / Name / Team / Score per cost* / Score (estimation) + Show Spoiler + 1 (Z Alone SKT 46.00 46.00 --------- team* sorted by it A nice heuristic for spotting the proper picks is to notice that they form local optima in this sequence. For team you want the most expensive players who manage to stay cost-efficient, i.e. high on the list; for anti-team you want the cheapest players who manage to stay cost-inefficient, i.e. low on the list. In both cases the picks stand out from the crowd. P.S. I don't expect too good results from all this, because the gut feelings such as predicting based not only on last PL are actually pretty useful. But I shall stick to the approach anyway just to see how (bad?) it will end up. | ||
|
ShadeR
Australia7535 Posts
On November 25 2011 21:29 figq wrote: As a scientific experiment, I decided to toss aside my human insight. That includes the promising Killer, Bogus, Wooki, and returning Effort. Instead, I shall rely strictly on evidence. I built a complete list of FPL scores for all players, based on last Proleague. Then I divided these scores by their current FPL costs just to get a good intuition about cost-efficiency, and understand the results better. Finally just used 0-1 Knapsack to calculate the optimal 30-team and 13-antiteam with the original FPL scores (not divided). That's based on first setting the main team, to be the one with strongest top 5, because a good team is worth more than any one player with this scoring. The results from all this were not too surprising. In fact, I've considered very similar layout beforehand on my own, but it's nice to have it confirmed by impartial numerical evaluation. Will edit with the team itself, and probably some more comments, after the deadline. Anyway to adjust for the formatt change? I assumed you excluded winners league stats but Bo5 no ace is significant. | ||
|
figq
12519 Posts
On November 25 2011 21:32 ShadeR wrote: No, I'm not using any adjustment. But even this way, the ace players generally don't end up looking very cost-efficient.Anyway to adjust for the formatt change? I assumed you excluded winners league stats but Bo5 no ace is significant. (i guess they are slightly over-priced) | ||
|
blubbdavid
Switzerland2412 Posts
1. End up in the upper half of the ladder 2. Beat Torte de Lini | ||
|
lequinow
Canada17 Posts
| ||
|
TheRavensName
United States911 Posts
| ||
|
Lonyo
United Kingdom3884 Posts
On November 25 2011 22:14 lequinow wrote: I don't quite understand why so many people put effOrt in their team. He was good once, but after his brief retirement, don't you think it's kind of a gamble to take this 5 points player? Assuming he hits near top form, he's easily an 8 point player, and you can get him for only 5. Every player is a risk. Flash has been injured, Jaedong was without a team and hasn't been able to practice. Baby as well etc etc. There are various players hwo have gone through problems which make them just as risky as Effort, only Effort is cheap. | ||
|
Subversive
Australia2229 Posts
On November 25 2011 21:22 Ikonn wrote: I can't believe the amount of people who put Action on their anti-team. As the #1 Action apologist on TL, I promise he'll put you all through a world of pain via plagues and swarms I can. He went 8-12 after winners league. And for 5 points he's a great anti pick. Means you don't have to end up with anyone 7+. On November 25 2011 19:53 Sawamura wrote: I can't believe people are putting kal on anti , I believe you are going to regret that decision soon, the gojila doesn't just fade out , he comes and back smash people anti team . Ace Goojila I believe and ace great !This I agree with. Kal was very unimpressive finishing 5-12 last 2 rounds. However, he's a good player and could come back to form. Plus he'll be on Ace and getting a lot of games. I'd much rather gamble with Action than Kal any day. On November 25 2011 22:28 Lonyo wrote: Assuming he hits near top form, he's easily an 8 point player, and you can get him for only 5. Every player is a risk. Flash has been injured, Jaedong was without a team and hasn't been able to practice. Baby as well etc etc. There are various players hwo have gone through problems which make them just as risky as Effort, only Effort is cheap. Yeah totally agree. Effort is a risk, but the bigger worry I have is that he's on a deeeep team. I assume he'll be at least as good as other 5 pointers (eg Mind, Roro, Action, Calm). And he has the potential to be worth 6,7 or 8 points if he's back in form. | ||
|
FaCE_1
Canada6173 Posts
lol | ||
|
Itachii
Poland12466 Posts
On November 25 2011 22:48 FaCE_1 wrote: I can't wait for it to start so I can stop changing my team T___T lol I changed mine only 2 times, and won't do it anymore.It always ends up badly, same goes with trading. Notraders unite! | ||
|
Shuray
Brazil642 Posts
| ||
|
Desert1fox
United Kingdom10 Posts
| ||
|
VGhost
United States3615 Posts
On November 25 2011 14:39 Crunchums wrote: I've always thought it was dumb because a) It makes the order of your wins matter (why should WWWLLL be more valuable than WLWLWL) b) It punishes your for the arbitrary circumstances of scheduling (a player with a schedule that alternates between playing good and bad teams is less valuable than that same player with a schedule that clusters bad teams together) and now it makes even less sense now that ACE matches are gone Actually you make a lot of good points. Was the original FPL a WL round? Streak points make more sense there. I still don't mind streak points, though to balance things if you're going to have them there ought to be - streak negatives to balance it out. Eh, whatever. | ||
|
amazingxkcd
GRAND OLD AMERICA16375 Posts
| ||
|
Tufas
Austria2259 Posts
On November 25 2011 21:13 Tufas wrote: I need your help Should I go 30/30 with great missing (with him I would have all 5 main ACE players) or take 28/30 with great (all 5 together are 18 points) Should I try to make a lot of points (well I have 4 ACE players anyway ..) or go for the all in with great Poll: Should I let great into my main team even if that means 28/30 points ? YES, as a ACE fanboy you have to do this. (13) No, 2 points wasted are really not in the spirit of the FPL. (6) 19 total votes Your vote: Should I let great into my main team even if that means 28/30 points ? (Vote): YES, as a ACE fanboy you have to do this. As you wish, great is on the team. | ||
| ||
