Looking for feedback on the current state of racing / Season 8

Just going to add here that another problem with your argument is that the lower participation in no way makes up for the level-difference.

Just look at shadow (10.4%) and compare it to marauder (11.9%).

Are you really going to sit there with a straight face and tell me that a 1.5% difference in participation makes up for 3 levels, or in other words shadow not having hit max bracket even once and marauder having hit max bracket dozens of times?

Fuck that.

My original statement was that it is (or was) generally possible to make a choice between more points or a better chance at 1st in class, and that this was fine and didnt need changing. It was a response to jstqs comment about 1st in class not being indicitave of player skill and that the system should be changed, not an attempt to disprove his statement or get into a stats arguement.

Anyway, aside from those examples with a sample size of one event, youve put more effort into digging up stats than im willing to, so i dont really have much more to say as far as stats go, and im not really convinced most stats are hugely relevant in the first place considering the race population is so tiny at this point.

"

KoTao wrote:

My original statement was that it is (or was) generally possible to make a choice between more points or a better chance at 1st in class, and that this was fine and didnt need changing. It was a response to jstqs comment about 1st in class not being indicitave of player skill and that the system should be changed, not an attempt to disprove his statement or get into a stats arguement.

Anyway, aside from those examples with a sample size of one event, youve put more effort into digging up stats than im willing to, so i dont really have much more to say as far as stats go, and im not really convinced most stats are hugely relevant in the first place considering the race population is so tiny at this point.

That's a great way of saying "I don't have any evidence to back up my claim".

The most recent couple posts got me thinking about what makes a class "powerful"

I don't think going off the highest individual in class or the sig record is a fair measure of how good a class is. It could after all just mean that the top players of that class are better than the top players of other classes.

To take a look at this more objectively I think we need to look at the top of the ladder and determine which classes are represented more strongly. Since each class has a different level of participation Im sticking mainly to percentages here.

I wondered to myself: what percent of witches make it to the top 100, compared to the percent of duelists that accomplish the same? Is this data interesting? Would others find this interesting?

For the purposes of discussion I have composed a couple graphs of two descent champtions events. There's a lot of factors that this data doesn't take into account so don't read too much into this (as you shouldn't with any out of context statistic you didn't compute yourself).

Disclaimer: I don't guarantee these stats are 100% accurate but from a spot check of the ladder they seem like a pretty fair assessment. I reserve the right to edit this post as I deem necessary

"

CreativeAspect wrote:

The most recent couple posts got me thinking about what makes a class "powerful"

I don't think going off the highest individual in class or the sig record is a fair measure of how good a class is. It could after all just mean that the top players of that class are better than the top players of other classes.

To take a look at this more objectively I think we need to look at the top of the ladder and determine which classes are represented more strongly. Since each class has a different level of participation Im sticking mainly to percentages here.

I wondered to myself: what percent of witches make it to the top 100, compared to the percent of duelists that accomplish the same? Is this data interesting? Would others find this interesting?

For the purposes of discussion I have composed a couple graphs of two descent champtions events. There's a lot of factors that this data doesn't take into account so don't read too much into this (as you shouldn't with any out of context statistic you didn't compute yourself).

Disclaimer: I don't guarantee these stats are 100% accurate but from a spot check of the ladder they seem like a pretty fair assessment. I reserve the right to edit this post as I deem necessary

Spoiler

Sorry these turned out larger than I hoped.
Image one
Image two

Some of these columns are labeled very well - % of pop refers to the number playing x class over the number playing.

An example interpreted in words:
In season 8 event 90, 6.8% of the top witches made it into the top 50. only the top 1.6% of duelists did this.

Well it's fairly easy to point the flaws in your argument. I mean given your own data you would have to come the conclusion that ranger is the third most powerful class, even though the top ranger record is 4 levels below the top witch record and 3 levels below the top marauder record.

What does that tell you? Maybe that the percentage of people per class in the top50/100 is totally meaningless.

Because while it is true that 3 rangers made it into the top 50 (one of which being myself), none of them will EVER make it into the top 20 or top 10.

I can maybe see a ranger hit 24 at some point with a really good run, who knows maybe even 25, but 27 like witches and templar or 26 like marauders?

No way.

There is no doubt on my mind that had I played any other class other than ranger (with the exception of shadow) I would have placed higher overall. No doubt at all.

I mean I did this before, for the longest time my personal best in vanilla descent on ranger was lvl 23.x, then I did one vanilla descent as ground slam templar and hit 24... first try.

It actually wasn't until recently (just a couple days ago) when Jani and myself beat boofs old vanilla descent ranger record and both hit 25.

For the longest time my personal best in vanilla descent was with a class I had only played once in that race. How stupid is that?

"

SlixSC wrote:

Well it's fairly easy to point the flaws in your argument.

"

CreativeAspect wrote:

There's a lot of factors that this data doesn't take into account so don't read too much into this (as you shouldn't with any out of context statistic you didn't compute yourself).

I'm not trying to make an argument (other than some basic conclusions about the potentially unreliable nature of statistics) and by my own words I'm admitting that two examples isn't enough to prove anything, I just think looking at empirical data might help paint a more accurate picture of the current state of racing. There is no singular end-all be-all that can summarize the entirety of POE racing.

"

SlixSC wrote:

I mean given your own data you would have to come the conclusion that ranger is the third most powerful class.

What does that tell you? Maybe that the percentage of people per class in the top50/100 is totally meaningless.

If you only looked at one graph and completely ignored the second you might consider that as an option. (in regards to ranger as third most powerful) However taking the second into account blows that theory away and would indicate that in regards to the strength of a ranger relative to other clases, this data is meaningless as it doesn't provide a consistent result (and I'm totally ok with that).

I think the only thing the data does indicates is that witch and templar have at least some advantage relative to the other classes. However with a sample size of two, this clearly doesn't prove anything other than indicating that a trend may exist.

"

SlixSC wrote:

There is no doubt on my mind that had I played any other class other than ranger (with the exception of shadow) I would have placed higher overall. No doubt at all.

The data supports this as a possibility, more strongly in regards to switching to templar or marauder, but not necessarily the other three classes.

"

SlixSC wrote:

For the longest time my personal best in vanilla descent was with a class I had only played once in that race. How stupid is that?

Great question. Is there a problem here, and if so what can be done about it?

I acknowledge that 100% class equality is likely impossible, and please keep in mind that this graph might look completely different when we look at a different type of race. But does the current scenario (at least in regards to D:C as a signature event) fall within the acceptable limits?

"

SlixSC wrote:

That's a great way of saying "I don't have any evidence to back up my claim".

Its a great way of saying "Bickering over stats and technicalities isnt going to convince ggg to improve racing no matter who imagines theyve won the arguement, and i wasnt particularly interested in such an arguement in the first place".

"

CreativeAspect wrote:

Great question. Is there a problem here, and if so what can be done about it?, more strongly in regards to switching to templar or marauder, but not necessarily the

I acknowledge that 100% class equality is likely impossible, and please keep in mind that this graph might look completely different when we look at a different type of race. But does the current scenario (at least in regards to D:C as a signature event) fall within the acceptable limits?

If how well you can do depends more on the class you pick as opposed to how good you are playing the different classes then yes I see a problem there. I don't even see why we are discussing this now, I mean it's obvious.

"

Its a great way of saying "Bickering over stats and technicalities isnt going to convince ggg to improve racing no matter who imagines theyve won the arguement, and i wasnt particularly interested in such an arguement in the first place".

LOL, just fucking lol.

If you are making an empirical claim like "Playing an underpowered/underrepresented class is a fine choice too, as the potential increased demi chance is offset by the generally decreased points due to lower final level." dont cry foul when someone points out that the data doesn't really support your argument.

Number of Flameblasts for Shadow: 1

Prediction on the next Flameblast drop: S08S0??

Zoroxo plz :(

"

SlixSC wrote:

LOL, just fucking lol.

If you are making an empirical claim like "Playing an underpowered/underrepresented class is a fine choice too, as the potential increased demi chance is offset by the generally decreased points due to lower final level." dont cry foul when someone points out that the data doesn't really support your argument.

Its been a common way to take easy demis since closed beta. Your cherry picking pieces of individual seasons or specific events to show otherwise doesnt change this.

Youll note that jstq mentioned doing the exact same thing in the statement i originally responded to, btw.