Hi, well i was talking to janmb yesterday and agreeing that the turn success and safety success tables on killer are very unfair, as people just simply stop once they reach 500 turns.

If you look on the safety table, there is only 1 player in the top 10 that has barely passed 1,000 turns.

Personally, i think that either the turns required should be upped to 1,000. Or after a certain time of a user not playing killer, they get took off the leaderboard, as they are simply "retiring" at the top.

Cheers,

Damien.

Good point.

I just looked at the tables now, and my pot success is 7,192 of 9,202 (78.2%) which isn't the best, but considering it's thousands more pots than most people on there, it seems more significant.

Although no disrespect to anyone who is actually on that table.

I agree.

No disrespect to anyone, we all play according to the system the way it is, the question is as always whether that system can be changed for the better.

It's currently possible to play 500 shots, reset, play another 500 and so on and so forth until you get percentages you are happy with. Once a good score has been established, you can shut out that table indefinitely.

Those tables simply need to diminish over time, much like rank does, so it's not possible to sit on an accomplishment forever. I know it can't be compared to rank, but the problem is the same in principle.


So, to bring suggested solutions and not just complain, lets look at how it could work...

1: Keep scores as-is for one week after the most recent game played. After that, the score (on the tables only, not on the users profile) start diminish by a given amount once a day (the value having to be a result of study and tuning)

2: Leave the system as-is, but make the minimum number of shots required dynamic instead... After reaching 500 shots, require a given number of extra shots per week (again subject to tuning) in order to still qualify to get the score on the tables.

3: Introduce an alternative sorting, not using only the percentage itself, but also factor in the number of games played (for the win percentages) and shots taken for the pot success/safety percentages.


Both solutions 2 and 3 are simpler and easier to implement than the first.

Edited at 13:05 Mon 25/05/09 (BST)

I have not set any targets with regards to killer for myself but i have noticed this and there are players who take advantage of it so i completely agree with what damien and janmb are saying.

Also i favour solution 3

I totally agree even though i have not had 500 turns i still think we have to have 100games played to not be a newbie but if it is based on turns then it should be alot more not just than 100 turns but 500 turn so therefore it is more fair for a person to be at the top of the table but whom as played about the same amount of games as everybody, therefore i think it sould be atleast 1,000 turns before they make it onto the table.

I have some sympathy with the views expressed here having sat in the top 10 of the straight potting percentage table for ages with my percentage based on approx 30,000 shots, which was anything from 7 to 60 times the number of shots played compared to those above me. Simply upping the number of qualifying turns wouldn't end the problem of people retiring at the top.

The problem with statistics is that they can be massaged to produce whatever result you want them to - that's why governments use them so well! By definition a pot success rate or a safety success rate can only be calculated the way they are now. Using any other factors to influence the outcome is really just proving the above fact. For example I play straight friendly games with a whole different approach to straight competitive games so either including all types, or not, would also give a different outcome.

Options 1 and 3 simply don't give a success rate although option 3 could give you a completely new stat if its what you really wanted. In the grand scheme of things though is formulating a completely new stat really all that important?

On that basis only option 2 comes anywhere near being acceptable if you say that accounts become 'inactive' for the purposes of these tables if they haven't played a set amount in a set time. But that still doesn't stop anyone making it to the top of the table in their first 500 shots again and again, either through resetting or through new accounts.

In fact you could argue, and I am sure spinner would, that indeed why should it, as getting to the top of these tables is as much of a challenge, and a valid way of playing the game, as it is for those who wish to gather points towards getting a high score in the 'ranking' table.

I agree that anything that modifies the percentages, dynamically or not, changes what they are. I guess what I'm asking for is just that: A new stat to replace the current percentage tables.

Also note that we're talking about the tables here, which does not have to affect the stats you keep for each user on their profiles.





Well no, they could still do that, but they would drop after a short time, OR do what we want them to do: Having to play more games on top of their initial scores.

Well no, they could still do that, but they would drop after a short time

if they drop - then they no longer reflect a potting or success rate and are therefore meaningless

only making them inactive, through whatever qualifying means, - ie removed from the tables not deactivated - achieves what you all want.

Precisely.

You misread my last post tho - in that case I meant they would drop, as in the users would be dropped from the tables. Not talking about dropping scores over time in that post.

ahhh ok then, my apologies

I still think you will get a constant stream of new accounts flying through at the top so it might not improve things a great deal but it certainly does no harm either.

Adding a growing requirement to played games and shot counts would probably be better since it leaves the numbers intact and as they should be, while avoiding the reset problems.