ZakStunts 8.0

[CTG]

Whohoooooo, new perspectives in useless stats: personal car preferences, replay distribution in a typical month, ranking vs number of sent replays correlation, etc.

[dreadnaut]

Minor issue: in some seasons, the first race started during the final days of December (Z59 in 2006 and Z91 in 2009). In such cases, said race is being split between the two seasons.

Darn, my smart way to detect seasons isn't smart enough

[Duplode]

Minor issue: in some seasons, the first race started during the final days of December (Z59 in 2006 and Z91 in 2009). In such cases, said race is being split between the two seasons.

Darn, my smart way to detect seasons isn't smart enough

Last time I played with ZakStunts historic data the replay entries carried season information...

---

Another little thing: I really appreciate that there is a link to Southern Cross in the Downloads page, even though there is a typo in the link 

[Duplode]

And another: the season scoreboard is still calculating SUM according to 2012 rules (i.e. dropping 4 rather than 3 races).

[dreadnaut]

Last time I played with ZakStunts historic data the replay entries carried season information...

They do, great! Stats now show the correct tracks in each season. Thanks Duplode!

Another little thing: I really appreciate that there is a link to Southern Cross in the Downloads page, even though there is a typo in the link 

Fixed!

[Duplode]

Two more details:

• Most internal links in the site are relative, so they break when scoreboard pages are accessed through the  pretty URLs used for the automatic links in the shoutbox, such as http://zak.stunts.hu/tracks/ZCT146. (On the other hand, pretty URLs would fix named anchor links such as the the stressed article.)
• Markdown formatting is not being applied to "official" posts in the news archive.

[dreadnaut]

Markdown support is still incomplete, also because I'm not sure if Zak is interested in using it —Zak, are you interested?

Links: yep, all broken and they are quite a few. I can 'fix' that temporarily with <base> tag, but I need to change them all over the place, if I keep the pretty urls. However, there's a bit of a feature queue at the moment, all stuck behind actual work

[CTG]

A strange mistake: check satanziege's performance for 2013. Two races are missing from his stats - however, the sum score is correct.

[alanrotoi]

He is desapearing from the reality... maybe he is now a...

[Duplode]

A strange mistake: check satanziege's performance for 2013. Two races are missing from his stats - however, the sum score is correct.

Seems that results below 12th are not being displayed in the performance chart.

[alanrotoi]

I have an idea about some rules, help me to complete it. I was watching the season's standings and I compared Aburaf's and mine results. Anelio has a SUM = 68 and REAL = 56. I have SUM = 59 and REAL = 52. I guess this rule prices the lazy pipsqueaks like me and gives a disadvantage to regular pipsqueaks like Anelio.

What about for the next year if we change the rule "three worst results don't count at the season's rankings" to something like "two or three missing races doesn't count at the season's rankings" or something more fair in this case?

I mean this rule took 9 aburaf's points and to me only 2 points. Maybe this rule is more useful when there are more pipsqueaks because now if I race this month I'll have 1-4 points not because my performance but the lack of pipsqueaks.

[CTG]

I guess this rule prices the lazy pipsqueaks like me and gives a disadvantage to regular pipsqueaks like Anelio.

Regular pipsqueaks have endurance bonus.

But let's compare the results of you two (finished races, after ZCT147):

ABU: 10-9-7-6-6-5-5-4-4-4; EB: +1; sum: 61, minus the worst three results or missed races (4-0-0): 57 pts
ROT:   9-9-8-7-7-5-4-4-3-2; EB: +1; sum: 59, minus the worst three results or missed races (2-0-0): 57 pts

Let's presume that AbuRaf will stay on the 6th place, earning 7 points with that - while you miss this race (hopefully not).

ABU: 67 pts from races; EB: +3; sum: 70, minus the worst three results (4-4-0): 62 pts
ROT: 58 pts from races; EB: +1; sum: 59, minus the worst three results (2-0-0): 57 pts

I think it's a decent difference. Now let's go further: AbuRaf finishes 7th in December (average ranking for him in this season), while you: 1, miss the race; 2, finish between 1st and 12th.

ABU: 73 pts from races (6.08 pts/race); EB: +6; sum: 79, minus the worst three results (4-4-4): 67 pts
ROT, first scenario: 57 pts (5.8 pts/race)
ROT, second scenario: 59-70 pts from races; EB: +3; sum: 62-73, minus the worst three results...
- 12th place: 59+3-(2+1+0)= 59 pts (5.36 pts/race, from pure race points)
- 11th place: 60+3-(2+2+0)= 59 pts (5.45)
- 10th place: 61+3-(3+2+0)= 59 pts (5.55)
- 9th place: 62+3-(3+2+0)= 60 pts (5.64)
- 8th place: 61 pts (5.73)
- 7th place: 62 pts (5.82)
- 6th place: 63 pts (5.91)
- 5th place: 64 pts (6.00)
- 4th place: 65 pts (6.09), slightly better average ranking than AbuRaf's, still behind him.
- 3rd place: 66 pts (6.18), significantly better average ranking, still behind him.
- 2nd place: 67 pts (6.27), draw
- 1st place (very unlikely, but still, for the complete table ): 68 pts (6.36)

I think this system is fair enough with the current EB compensation.

[CTG]

For another oversimplfied example, we have two pipsqueaks: Stable Jim and Hectic John.

Stable Jim completes all the races, of course with a narrow ranking distribution:
4th - twice
5th - four times
6th - four times
7th - twice

According to the current system, he would have 90 pts from races, +6 from EB, -19 from the worst three races ---> SUM: 77 pts. His average ranking was 5.5.

Hectic John is an unpredictable pipsqueak, his performance (and participation rate) chart looks like a mad rollercoaster track.
1st - 2 races
2nd - 2 races
4th - 1 race
5th - 1 race
7th - 1 race
9th - 1 race
10th - 1 race
missing: 3 races

That's 76 pts from the races, no bonus/reduction. Average ranking: 4.56 - although it's a lot better than Jim's (containing even 4 podium places), he lost the duel because of missing three races.

[CTG]

Will we ever see ZakStunts 9.0?

[zaqrack]

As we practically have new code under the hood at several places, I have updated version to 9.0