Re-Organizing DSB

[Entr0py]

I asked if these are practice stats, or if league is included.

The stats I gave earlier included all pracs and leagues played on my bot since 2010-03-11 08:24:35. I do not have the old DSBL stats integrated with the bot or website, and so they are not included in the calculation.

I'm sorry but your idea that there are more kills in any other sector than in West, is a completely fucked and wrong idea.

...

I'm telling you west, doesn't see less kills than other sectors. I would argue you see even more kills from west players.

...

SECTORS DON'T AFFECT KILLS.

...

A SECTOR DOES NOT EFFECT HOW MANY KILLS A PERSON MAKES.

And how often players kill depends on the sector.

WRONG, FUCKING WRONG, RETARDED. WRONG, STUPID, WRONG.

IT MIGHT - MIGHT - HAVE A SMALL - SMALL - AFFECT ON PEOPLE - ON PEOPLE NOT KILLS! ON PEOPLE. THUS MAKING THEM MORE LIKELY TO FOLLOW A CERTAIN STRATEGY. BUT THIS IS AN INDIVIDUAL CHOICE AND USUALLY PEOPLE PLAY HOW THEY WANT TO PLAY NO MATTER WHERE THEY ARE. EVEN IF IT MEANS LOSING. THAT'S WHY LOTS OF PEOPLE SUCK ASS AND OTHERS ARE NATURALS. The only real strategy is get the fucking flag at the bell, kill the bitch in your way, lock out the sector.

You are just wrong. Here are the stats for each sector, broken down by league, as well as the overall stats rolled up for each sector. We can clearly see that consistently, west has the least number of kills of any sector. Furthermore, there is a significant difference in the number of kills from one sector to another.

Code: Select all

Overall Statistics by Sector:
Sector                  # Player Match Entries     # Matches     Average Kills Per Player (Normalized to 20 minute game)
-----------------------------------------------------------------------------------------------------------
East                    49627                      7516          21.13865910
North                   11586                      1384          20.16467189
South                   2596                       388           18.44031889
West                    1173                       261           16.77799676


East Statistics by League:
League                  # Player Match Entries     # Matches     Average Kills Per Player (Normalized to 20 minute game)
------------------------------------------------------------------------------------------------------------------------
Indi League 1           41086                      6255          21.39348225
Sector League 1         53                         9             22.18525283
Setting Test 1          18                         3             21.62612778
Sever System Test 1     58                         10            22.28936897
DSBL 14                 201                        16            17.37294776
Prac Sector League      2429                       370           20.51025414
Team Combat League 2    2                          1             34.1969
Prac Sector League 2    5780                       852           19.69515528


North Statistics by League:
League                  # Player Match Entries     # Matches     Average Kills Per Player (Normalized to 20 minute game)
------------------------------------------------------------------------------------------------------------------------	
Indi League 1           9815                       1150          20.33700584
Sector League 1         17                         2             20.91660588
Sever System Test 1     9                          1             19.50105556
Prac Sector League      594                        70            18.63232172
Prac Sector League 2    1151                       161           19.48000582


South Statistics by League:
League                  # Player Match Entries     # Matches     Average Kills Per Player (Normalized to 20 minute game)
------------------------------------------------------------------------------------------------------------------------	
Indi League 1           2092                       315           18.35876319
Sever System Test 1     6                          1             18.5
Prac Sector League      137                        21            17.80463577
Prac Sector League 2    361                        51            19.15318892


West Statistics by League:
League                  # Player Match Entries     # Matches     Average Kills Per Player (Normalized to 20 minute game)
------------------------------------------------------------------------------------------------------------------------
Indi League 1           1006                       230           16.60029254
Sector League 1         18                         4             14.25703889
Sever System Test 1     8                          2             18.875
Prac Sector League      50                         10            15.443638
Prac Sector League 2    91                         15            19.78996813

PS - If you've been playing west 3v3 lately, of course it will produce different results from earlier times. That's a totally different game from 2v2 west.

PPS - For the east stats, I'd disregard the DSBL14 and TCL2 lines. It's clear that the bot wasn't set up correctly for that, and flagged those game types as being played in "East".

[falconeer]

In 2010 we didn't have any 12 vs 12 leagues except DSBL 12-13? Any DSB league after league 10 was in serious decline. Down from 400 players to 100. Your stats are based solely on practices with your bot and sector league.

Look at the west games played, 1000 vs 49,000 east. How about you dig up a game that has the most amount of kills in west. It's probably 35+ kills. Tied with the most amount of kills in East. These stats are still bogus. You need the same amount of games in each sector to run proper comparison averages. =/ This is the average of 1000 games west, and 49,000 games east. They are accurate, but not comparable.

But I love to be proven wrong. In all seriousness, I don't ever like being right. If I'm right it's bad, because I'm cynical, and it means other people are wrong. I'd rather be wrong about you being wrong. So +1 for proving some of what you are saying.

I mean judging by the stats and player entries. The more player entries you have, the higher the kill averages are. It's noticeable.
49,000 entries, 21 avg east
11,000 entries, 20 avg north
2500 entries, 18 avg south
1000 entries, 16 avg west

Why would North lose to east when RN is as fierce as east, and there are more players playing north? Its because east has 49k entries, and north got almost the same average in just 11k entries. For all I know if you include another 30k practices in north, the average will jump up to 22 kills on average and beat east.

So how many of those 1000 west entries are from DSB league 12-13? Like 50? The rest are practices. This is bad. Compare league 12 and 13 sector stats with each other only, and you might get totally different results. Like accurate results for one. Going further compare league 1-10 stats and you'll get different kill averages that reflect better what happens in each sector.

But I doubt even league 12-13 were included in your stats, even though they happened in/after 2010.

[Entr0py]

The only leagues included in the stats are those listed. DSBL12/13 must have been played on a different bot, and so their stats were not captured by this bot.

You *might* be right that with 30k more north/south/west matches, that they would all converge on the same number of average kills as the east average. But the chances of that happening are astronomically low, considering there are 100+ sample points for each of them, and thousands in the case of north.

[falconeer]

How would you go about collecting stats from all the previous leagues? Is there any menial repetitive work I could do to help achieve that goal for you? Because I'm curious what all the 12 vs 12 league stats would yield out. Don't we have these stats on Frankillers stat website? Or on league.nu?

Plus in your stats, I'm still curious what the most amount of kills were in west, in a game. It was probably 35+ kills. Right? So just because it doesn't happen often, doesn't mean it doesn't happen. For whatever reason it didn't happen often like it did in east. Either shit players, or people preferred to run around flagging like freespirit instead of kill etc. The sector doesn't dictate how many kills a person can get. =/

[Entr0py]

How would you go about collecting stats from all the previous leagues? Is there any menial repetitive work I could do to help achieve that goal for you? Because I'm curious what all the 12 vs 12 league stats would yield out. Don't we have these stats on Frankillers stat website? Or on league.nu?

Plus in your stats, I'm still curious what the most amount of kills were in west, in a game. It was probably 35+ kills. Right? So just because it doesn't happen often, doesn't mean it doesn't happen. For whatever reason it didn't happen often like it did in east. Either shit players, or people preferred to run around flagging like freespirit instead of kill etc. The sector doesn't dictate how many kills a person can get. =/

The most ever recorded on my bot was 44 kills in 1200 seconds by 45c on 2010-03-14 01:53:41.

As for collecting stats from previous leagues and putting it into the website database, it's possible, but it would be a challenge. I'm not entirely sure how to go about it just now, I'll get back to you on that.

[jim the chin]

If we want kill points equal to flag points, we should use:
East: K = 70.96 * 1 = 70.96
North: K = 119.05 * 1 = 119.05
South: K = 97.61 * 1 = 97.61
West K = (2 players): 107.27 * 1 = 107.27
West (3 players): K = 107.27 * 1 = 107.27

Ent, do you have any objection to using this from your first post, which doesn't include the kill bonus. Then, we set K=95 for the arena, which is the midpoint of your values. Temb and me can both live with K=95. Then, we add a kill bonus that doesn't affect these numbers. As it's a flag-based and a kill-based quantity, it doesn't need to go into your previous equation, and it doesn't need to affect the numbers quoted above. Instead, we could try this for the kill-bonus:

Points from kill bonus = 0.4 * Points from bell flags.
Which is:
Total kills * B * number of flags / 2 = 0.4 * Points from bell flags.
Where: `Total Kills = avg kills per player * number of players'.

The factor `0.4' is arbitrary, and it means I'm making the kill bonus worth 40% of what bell flags are worth in the game. We want to work out B for the different sectors, before picking an average value of B as the arena value.

EAST: 21.139 * 6 * B * 20 / 2 = 0.4 * 9000
NORTH: 20.165 * 8 * B * 32 / 2 = 0.4 * 19200
SOUTH: 18.440 * 6 * B * 24 / 2 = 0.4 * 10800
WEST(2): 16.778 * 4 * B * 24 / 2 = 0.4 * 7200
WEST(3): 16.778 * 6 * B * 24 / 2 = 0.4 * 10800

East: B = 2.838
North: B = 2.975
South: B = 3.254
West(2): B = 3.576
West(3): B = 3.576

Without needing to do any averaging, it's clear that B = 3 is ideal. Given that this is for when kill bonus points = 0.4 * flag points, we end up with:

Kill bonus = 16.7% of total points.
Flag points = 41.65% of total points.
Kill points = 41.65% of total points.

---------------------

So this is what I propose:

Kill: 95 (reduced from 150).
Kill bonus: 3 * flags held. (reduced from 5 * flags held).
Bell flag: 10 per flag per player (same as it is now).

Is this agreeable to everyone for now?

[Tembest]

Hard to agree on people insisting to change something that worked.
All of sudden it was all about finding the balance between the total distribution between kills and flags, dsb has always been a flagging game, and now for
the first time we are going to change the emphasis. I guess it's not so bad. It's just funny how it turned into average kills and all that information that was
never the topic. 6vs6 league worked, and the scoring system was praised by people. The ratio was 1 kill equals 2.5 bell flags. Bell flags still contributed
more to the total points than kills.

Anyhow, I'm not here to argue, because what Sever proposed is still 1000 miles better than what he's been proposing for the last week.
I would still rather set the kill to 90 than 95. It does not make a major difference but 90 is simply easier to follow than 95, and 100 would exceed suggestions
(except Sever's) clearly.

6v6 was the last league with a functioning scoring system. Now we are making calculations and all. I just cannot understand why don't we aim as close as
possible to something that worked. That is what is confusing me. Everybody (including me) claims to know what is good for the game and all. We look at
crazy averages and make silly assumptions how points will be distributed etc. etc. The gameplay will change the second we change the scoring system and
people adapt to the new system. This is actually a serious real life problem as well, developers, mathematicians make assumptions based on historical data
and act surprised when people change their behaviour after new implementations... so many times the stupid guy on the field who can barely start a laptop
is so much better at evaluating human behaviour and consequenes of changes than statistical analyses.

[jim the chin]

FFS... Well lets go with 90 then.

Anyhow, I'm not here to argue, because what Sever proposed is still 1000 miles better than what he's been proposing for the last week.

Hmmph, ok. I change my mind because, 1. I want to reach an agreement that works for everyone; 2. I'm not religiously sticking to one theory based on one limited scenario that certainly doesn't define the entirety of the game; 3. The more I read, the more I learn. I don't ignore the things that disagree with my preconceptions; and 4. I try not to act like this is some kind of personal war.

and 100 would exceed suggestions (except Sever's) clearly.

Uhuh, anyway, I hope we can all agree on K=90 and B=3 for the changes, and I hope north is still fun to play after this.

I'd still like to see the old discussion about this topic if you have the link.

[falconeer]

I have to post for snake because he's given up on you guys at this moment lol.

[10:11:00 AM] Ziv: how many times ppl were mvps with most kills in south/west
[10:11:02 AM] Ziv: back in the days
[10:11:04 AM] Ziv: its bullshit
[10:11:15 AM] Ziv: and someties u hardly get kills in east coz everyone is hunkering down

Also, I still agree with Tembest, but 90 is a good compromise.

Also, the old discussion is long gone, the forums are gone, maybe the way back machine has it. But all I can say is that it was huge.

[jim the chin]

[10:11:00 AM] Ziv: how many times ppl were mvps with most kills in south/west
[10:11:02 AM] Ziv: back in the days
[10:11:04 AM] Ziv: its bullshit
[10:11:15 AM] Ziv: and someties u hardly get kills in east coz everyone is hunkering down

I'm sure snake knows what averages are.

90 is a good compromise.

Good.

Also, the old discussion is long gone, the forums are gone, maybe the way back machine has it. But all I can say is that it was huge.

Temb, said he read it recently:

I bothered to look up the old discussion. I proposed anything between 2-3.