iDB Community Forum

Hi

I must thank to all the iDBers involved on the RoS data gathering and the organizers to allow me to grab that data and use it to practice on spss. Also if someone has a new idea i will welcome it (i need some practice).

Here some little findings.

If we could make an average battler according to this data it would look like this:
HP 451.55 (Min. 3 Max. 7,500); STR 292.94 (Min. 9 Max. 2,039) and DEF 195.57 (Min. 5 Max. 1,266).
And his score table would be:
Junior 466.85 (Max. 4,502); Comm 187.46 (Max. 2,302) and Riilan 65.52 (Max. 1,341)
Grundo 232.66 (Max. 3,160) and Militia 103.99 (Max. 1,835)
Bot 22.48 (Max. 2,000)
Score 37,695.03 (Max. 709,220)

Lower leagues: League 1 is doing better than league 2; also against Riilan it perform better than league 3 (L.1 has such low stats, then it has a bigger budget? noobs with their second battlepet or simply more experience).

Medium leagues: The upper echelons tend to fight less with the weaker opponents but, up to certain point, they are homogeneous.

Upper leagues: Almost all have reached the last boost on power but it looks like the difference is their defense but mostly their hp.

Now i will do some correlations to do more than describe.

NOTE: More below

edit: I will change the images to improve the accuracy when the data change (probably until the end).

Your upper-leagues image isn't showing.

So you're running averages? It's not really an entirely accurate system, in my opinion, due to weapons variance. Even though the upper-leagues have a wide HP variance, Defense (and a Tabard) means everything in this Plot.

Thanks Shattered... i fixed the image. You're right averages may not be the best thing but is just the start.

Below are some correlations.
Intuitively with bigger HP more victories and a bigger total score. Thats because one will have more opportunity to get more victories despite how lousy or amazing is their own battlegear.

Some ideas or different conclusions?

edit: I added some info.

One extra about participation. Almost two thirds of the participants are rank 9 or below. League 7 is the biggest, the second one is the 9th. The extremes (L.1 AND L.23) are the smallest.
Graph
Table

bulges at 5, 7, and 9, very intersting

I expected one at the limit of dubloon training, and another at the edge of ninja school, but not there...

"STR 292.94 (Min. 9 Max. 2,039)"
Why didn't you use the 700 boost limit?

The Min. & Max. are mostly just trivia stuff, and the decimals to show I have sense of humor (or to show my geekyness). I made some regressions with these raw data but I know the need to translate STR and DEF digits to boost measures. An example of what I can do is this (125-199)=3.0 but I'm not sure if that is meaningful to our purposes.

Some regressions below, proceed with caution (and tell me about any mistake). The total score is presented because it is an overall measure, the Victories try to gauge a direct impact of stats against a real challenge.

Total Score & Pet's Stats
Model Summary

Coefficients
Victories over Riilan & Pet's Stats
Model Summary
Coefficients
Victories over Grundobot & Pet's Stats
Model Summary
Coefficients
I would say that weapons as an independent variable to predict how much succes one will have on the battledome, at least on 1-p matches, becomes more important as the opponents difficulty rises. Some ideas or different conclusions?

Edit: More research later.

I'm sorry if this is slightly off-topic, but how do you calculate the total score for each individual participant, and also how often do you update your score page?

The score is calculated with the following formula: Score=[(X1 Opponent's Difficulty)*(number of wins)]+[(X2 Opponent's Difficulty)*(number of wins)]...

About the score page... I'm not sure if the team that keeps it updated has and schedule but I'm certain that A) they will says when the next update is due, or B) Say that the score table was updated. I suggest you to get as much victories as you can and forget the next update (toward the end check that your relevant data is correct).

This is probably party due to the fact that if you have bad weps you will max out more quickly of course. So then you'll go back and fight the weaker opponents more often, thus partly reducing the advantage in wins the stronger wepped people would otherwise have on those weak opponents.

::.Scraps an overly long analysis of One Player battling.::

Here's the problem: for a low-stat Pet, the net icon ratio continually shifts in your opponent's favor as the opponent becomes stronger. Thus you need more powerful weapons to keep that ratio "even" (or more specifically, in your favor). However, for a high-stat Pet (450 Boosts plus), the weapons matter less because that ratio is far more limited. A higher stat Pet with weaker weapons can perform (approximately) equally to a weaker stat Pet with stronger weapons.

Overall, yes, weapons are the most important thing, but it's because of stats.

Right, that's definitely true, I was just addressing the apparent case that weps become more important when battling stronger opponents.