Also, very few people are picking Zen because he is their passion deck. If a teklovossen makes day 2, the teklovossen player using a ton of skill they have picked up over multiple tournaments. Zen is just good and doesn't rely on 6+ months of testing to produce results.

@DiceCommando

Ай бұрын

100% (see what I did there - so witty). To be fair though, Zen is so new that it's hard to be anyone's "passion" - but could certainly have a ton of ninja mains who went over to him, to which that could count. But in terms of iteration and testing? Definitely not.

Really enjoy the variety of your content. Some deck tech. Some interviews. And then some commentary on the game itself including things like conversion rates. Cool stuff to fill the drive to work with.

@DiceCommando

29 күн бұрын

Well thank you, that's awesome to hear!

Good comments on comparing conversion rates and relative rates. It's a great point. I would really like to collect data on the players, ELO points, choice of hero, and placings. Which has greater predictability - the hero or the player? We could do this if we had the data. Simple statistical tests. Sure, I believe we would find that the players have more predictability, but how much do the heroes have compared to each other given their players' ELO points? In other words, can we take ELO points into consideration when rating the heroes? If we have the data, it can be done.

Conversion rates are always tough to use in general. It also takes a lot of assumptions that people playing are all the same. Not specialists or anything like that. One specialist as the only one on their hero getting top 8 makes a weird conversion rate for example. Also, tough when a really large group play the same hero because he can eat themself. Covered cannibalism 4:56

I think we might get more honesty and objectivity out of data driven arguments if we had access to more data and (especially) authoritative data. I can't easily tell you how many LL points Zen is earning from week to week and honestly compare that to Chane or Kayo. It's rare for me to have access to accurate tournament data that isn't being mined from streams highlights and isn't easy to compare from event to event. Access to this data would allow the data and stats people among us to visualize this data in all sorts of interesting and peer-reviewable ways.

Day 2 conversion is also an issue in mixed formats too. 7:09

2:30 thanks for covering the bad assumptions!

This is good shit, I've been driving this point too. Now we can talk about Schrödinger's Talishar statistics! One day they mean nothing because all players suck but on the other hand I have a pretty impressive WR with my deck in match up XX so maybe you should shut up unless you have stats to back your stuff up. Although if you do.. lol Talishar stats.

You articulate very well Andrew!

Maths is hard!!

Ah the annual DC conversion rate PSA. If we all play Zen, then Zen will have a 100% conversion rate, right? 😂😂 Speaking all of CC format: I wonder if some subset of the player base feels banning are necessary not because of some perceived dominance of a hero, but moreso a dominant meta archetype. I feel like some are more upset about the meta archetype being overwhelming aggro right now, but using Zen as the scapegoat because of “numbersssss.” It’s very young into MST meta. Set was released like 5-6 weeks ago at this point and decklists and meta are still developing. There’s just a ton of aggro right now because of the speed of the decks the perceived best way to deal with aggro is aggro them back. Right now there’s a lot of heroes who can perform the meta archetype well into other aggro decks - Zen, Kayo, Azalea, Dash (both), etc. To me, that’s pretty healthy. Others won’t agree and prefer to think of balanced metas with viability of all archetypes and not necessarily of variety of heroes. Also, card display between the framed mats when the light is bright is looking goooood.

I mean, aside from sample size everybody's, assuming there's a uniform underlying distribution and equal probabilities when players sit down, they're not independent, identically distributed, not across the heroes they play and not across the people that they play

How do you calculate conversation rait? Number of zen joined / number of zen in top 8? If you calculate Ratio of zen to other hero's/ ratio of zen in top 8. You get more real conversation rait.

@DiceCommando

Ай бұрын

I think what you're driving at is 'relative comparison' which I cover at around 09:40

There are lies. Damn lies and conversion rates- Ben Franklin

To counter your point: I don't think the commenter asking about usage rating thinks that just because you're running a deck like Zen means you're randomly going to get into top 8. It's more of Zen has more players, so has more chances of getting top 8 slots vs say 1 Tekklo player getting a top 8 slot. Both technically did well, but Zen is assumed to get more top 8 spots by virtue of there being more players. But say there are 7 Azalea players then all 7 Azalea gets top 8. Is it because the number of players playing Azalea, or is it because Azalea has tools to beat the current meta? That's when it becomes interesting

The amount of players playing a deck *absolutely* correlates with the amount of players in day2/top8. But yes, the rest of the video i agree with. Using a single value (e.g. conversion rate) to make a point based on Data is almost always wrong/misguided.

@DiceCommando

Ай бұрын

But it's because players gravitate towards the stronger heroes; it's not a product of merely quantity (rather quantity is a product of strength).

@Narabedla4

Ай бұрын

@@DiceCommando not merely. But it does correlate. Or would you say kano was a bad deck during PT1, because very few people played it? Similarly iyslander for a big part of her time was not played by a huge meta share, while being absolutely the best deck in the format for quite a bit. There are strong decks that dont get nearly the recognition they deserve by top players (one of them being Dash I/O that is still wildly underestimated with people thinking she gets fatigued by 3 blocks and mediocre fridge and you can hear that from good players in their respective tierlists.) But yes of course playrate is partly a result of a hero being strong, but it also double dips into top8 representation.

How dare you Andrew top players would never play a degenerate completely broken deck in order to win money and prizes. They have more scruples than that