Tbest, nobody said that, you said it yourself and then tried to prove it wrong, then you posted your "proof", that after 20 games, you'll not be rated more than 2000... and yet you have a rating of 2117, despite having played 21 games and having beaten not a single player ranked more than 2000... I mean, that has to be the funniest post ever.

What I actually said, was that runs will make your rating more volatile, and hence you have a higher chance of ending up with an inflated rating, so it is easier to get 1st (or a rating higher than you should) on a run. As you have just accidentally proved.

If your accidental proof wasn't enough, I'll prove it mathematically, lets say that your true rating is 2200, and to gain 1st place you need a rating of over 2300. After 1000 games, you should end up very close to 2200, there will be basically a 0% chance of being rated over 2300. After 20 games, you're rating will be less precise, you may be rated higher or lower than 2200 depending on luck, lets say there is a 10% chance of deviating from your rating by over a 100 points. so 5% chance of being rated above 2300, and 5% chance of being rated below 2100. And so a 5% chance of being rated 1st.

Therefore, there is a higher chance of being rated 1st if you played 20 games compared to 1000.

Edited 4/25/2016 17:44:33

LOl @ Mac'n'Cheese, Yes, someone did say "An issue with ladders is that its much easier to get #1 from runs than playing a lot of games over a long time." Guess who?

Anyway, since you didn't check facts here are a few. After 20-0 my rating was barely above 2k (~2015)
My 21 win was against a 2k+ (who lost on boot, and after the loss his rating is 1999)

Now, since you completely missed my point, let me reinstate it.

Currently it is harder to get 20-0 and grab 1st then before. (2-3 y. ago) Because 20 wins don't guarantee you high enough rating. In a theoretical view as the ladder grows (say 1k ranked players) the sample size you need is bigger then 20 to even theoretically have the chance at 1st. Thus theoretically you need a longer win streak, which in turn, theoretically would increase the chances of a loss.

Glad you found my post funny btw. Since it was written to be humorous. Through essentially I am making the same point that you are.

Edited 4/25/2016 19:21:17

Tbest, you only beat one player rated above 1850. You shouldnt be first with those opponents. I dont know why you got relatively weak opponents. I think it is still very much possible to get first because of . Ladserrun.

.

Edited 4/25/2016 19:42:18

It IS easier to get 1st from runs than playing lots of games, thats my point. Runs themselves are not easier than lots of games. If you do 100 runs and 100 lots of games, the average rating will be the same over all the runs compared to the lots of games, i.e. runs are not easier themselves, but the highest/lowest ratings will be runs, i.e. easier to get 1st with a run.

Yes, you're still rated 2117 after about 20 games, 21 and 20 are basically the same. You could have finished that game before one of your lower opponent games which you gained little from, and be rated around 2100 after 20 games. It was a boot win? this is what I mean, luck has given you a massive rating, despite only winning 20 games (and a boot), and you've not beaten anyone ranked above 1828. The luck evens out over a lot of games and so you dont get such an inflated rating.

I understood your point, I dont see how that "fixes the problem", My point is still true, that you are more likely to get 1st from a run than a large amount of games.

I also don't see how your point (runs are harder than 3 years ago) is essentially the same as my point (runs make it easier to get 1st than a large amount of games).

And I do believe it is possible for a 20 game player to take 1st, for example https://www.warlight.net/LadderGames?ID=0&LadderTeamID=12388 this person had a 2200 rating after 20 games despite 2 losses, if they had won those 2 games, they'd have easily taken 1st.

@Min, I never claimed that I am #1. And saying I only beat is 1 above 1850 is kinda misleading, since I won on boot xD Really it is more accurate to say that I have only beaten two 1800s. For my take on myself check out

https://www.warlight.net/Forum/148980-buns157-manipulating-1v1-ladder?Offset=0 (close to the bottom)

https://www.warlight.net/Forum/140471-managers-league-season-1-division-b-scores?Offset=120 (around the middle)


@Mac, I am saying you need longer runs, thus it requires more effort, and is harder to achieve ---> Less of a problem. Not saying, it is harder to do runs, then do Buns. Nope, Buns is way harder to do.

Also, https://www.warlight.net/LadderTeam?LadderTeamID=9019 has 2100++ after 6 games. SO yeah still possible, I am just saying that as the ladder grows in size it gets harder. In the past 2k+ usually gave 1st. Which is no longer the case. The general problem with runs is finishing 20 games before losing. For me 'accidental' run the problem is getting the rating up, not finishing enough games to get ranked. Reapiting my beloved slogan, Runs is still easier than doing Buns.

Edited 4/25/2016 20:30:52

But saying that you need a higher rating to take first makes runs harder makes no sense, it makes it harder for everyone to take first, including the people who have played a lot of games.

I don't know if that is your main or not, but if you joined in 2015, let me tell you runs used to be super easy. It was something you could just do if you were a top 10 player, and hope to do if you were a top 25, particularly if you had no problem with stalling.

Basically, the rating distribution and range was smaller, and the rating formula was the same. So ratings was relatively even more inflated, in terms of placing a player above an actual stronger player.

Edited 4/25/2016 20:29:02

Statistical discussions, yay.

What TBest states is true. A good run doesn't 100% guarantee taking 1st place on the ladder. But that doesn't mean that a 20-0 run doesn't give you an inflated rating. The standard deviation on a small sample size of games will always be high, as the standard deviation tends to decrease when the sample increases.
As the ladder grows, it is indeed harder to take a high rank with 20-0, though the effect of the ladder having an increase in players is relatively small.
Getting a high rating after 20 games is just as easy/hard as it was though. The general inflation of ratings in the ladder tend to make getting 1st place harder with a run (though the effect is relatively small), but the rating you'd get out of a 20-0 run should be the same as it was before.

So,
the fact that TBest didn't get a very high rating after 20-0 is due to him beating lower rated opponents than people do in their average run. Not due to the ladder having more players.
The fact that TBest's rating of 2110 isn't a top 5 rank, is due to the ladder having more players, and the ratings of the top 10 increasing as consequence.

Yeah, I joined in 2015, I've only really payed attention to the ladder. Even if runs were a bigger problem, they're still a problem, like farah says, the effect is quite small. And the ladder would have been WR then, which reinforces my original point, that WR exaggerates the power of runs.