Shooting Game Tournament 2008 Theme Weeks and possible Rules

[Shatterhand]

dont you have to die to abuse check points? sounds like it makes the game more interesting

I would refrain from saying something like that was interesting until I had played the game and tried to make a high score that way.

Viper Phase 1 thanks you.

[bsidwell]

Not Taito Week appears to have a Taito game in it.

[mikwuyma]

shhhh...don't destroy a shoddy excuse for a list of good games

[lgb]

No Hudson carivan? Aww.

[DMC]

mikwuyma asked me to explain DMC's scoring proposal since I explained it to him and he thinks most people don't get DMC's explanation. All he is suggesting is that the same scores are used for both individual and team.

Ex: an 8-man tourney with team A in the first 4 spots and team B in the bottom 4. Original proposal would be 6 + 5 + 4 vs 3 + 2 + 1. DMC suggests 8 + 7 + 6 vs 4 + 3 + 2, which are the same individual values anyway. Even simpler.

This is more fair and also gives the non top 3 finishers something to accomplish for their team. Don't worry, you cannot be penalized as a team for no-shows or terrible 7th members. The only way you are hurt is if someone's say, 4th place finishes above any of your top 3.

better expalanation indeed. One more thing though; maybe it's obvious, maybe not.

let the number of total registered players be the score for the person that wins a week. Not the number of participating players as that number changes from week to week. Otherwise one week will be worth more than another.

So if 146 players are registered:
1st - 146 points
2nd - 145 points
3rd - 144 points
...
146th - 1 point
and if people aint participating - just change their scores 0 points and keep their others as they are.

or if you want a fine looking number at the top just increase the points accordingly. e.g., 150 seems nice!
1st. - 150 points
2nd - 149 points
3rd - 148 points
...
146th - 5 points

edit: We had a similar contest in sweden and did this mistake, so first week gave the top scorer 53 points and the third only gave 39 lol.

[The Coop]

Amiga Week

Load up games like Agony, Project X, and Disposable Hero, sit back, and watch as the boards self destruct in an orgy of pissing and moaning

[Enhasa]

better expalanation indeed. One more thing though; maybe it's obvious, maybe not.

let the number of total registered players be the score for the person that wins a week. Not the number of participating players as that number changes from week to week. Otherwise one week will be worth more than another.

So if 146 players are registered:
1st - 146 points
2nd - 145 points
3rd - 144 points
...
146th - 1 point
and if people aint participating - just change their scores 0 points and keep their others as they are.

or if you want a fine looking number at the top just increase the points accordingly. e.g., 150 seems nice!
1st. - 150 points
2nd - 149 points
3rd - 148 points
...
146th - 5 points

edit: We had a similar contest in sweden and did this mistake, so first week gave the top scorer 53 points and the third only gave 39 lol.

We had a tourney at SDA and also had the same thing happen.

It's just a matter of preference. Some could argue that you're getting more points for beating more people. Since the only people the "mistake" affects are non-players for a given week (i.e. 1st player will still beat 2nd player by 1), it's not really a problem anywhere near the top. The old way allows someone not to play all weeks and still possibly get a decent overall score, while your suggestion rewards participation (even if it's just submitting a crap 1-1 score). Since I like participation, I agree with you again here.


Sidenote:

Mathematically this is identical to considering how to break ties. They are of course very rare in a shmup, but most people break them by giving tied players the upper end, e.g. 10, 9, 9, 9, 6, 5, etc. The old way is simply doing that for everyone that ties at 0 (no-shows get 1 point less than last place). The new way is giving all the non-players the lower end instead.

So the simplest way to explain scoring is: one point for each person you beat. In the above example: 10, 7, 7, 7, 6, 5, 4, 3, 2, 0, 0. (11 players in tourney, if you want top player to score 11, you can scale up by 1 but why bother)

[MR_Soren]

All weeks should be worth equal points.

[MX7]

All weeks should be worth equal points.

So you should get as many points for winning in a week where 150 compete as you do in a week where 15 people compete?

Hmmmmmm.

[MR_Soren]

So you should get as many points for winning in a week where 150 compete as you do in a week where 15 people compete?

Yes indeed! That's the only way to keep things fair. Here are some problems with basing points on weekly participation:


Participation wanes in the later weeks. Let's suppose the participation is as follows:

150
80
40
20
9

The players who do well week one have a huge advantage. Thus players good at week 1's game have an overall advantage.

Result: Contest is broken.


Even assuming less drop-off, a team that is in the lead could ask their weaker players to stop submitting scores, making it more difficult for other teams to catch up.

Result: Weaker players would feel unwelcome, which is against the spirit of the contest, IMO.



Now, I understand the problem with all weeks being equal. If each week is worth 150 points, and weeks one and two have 150 and 100 people respectively, that means you get 1 point for last place on week 1 and 51 points for last place on week 2.

Seems sad that you could get 50 points for submitting a terrible score just because nobody else played.

The fix: Low places are worth 1 each.

For example, start the score at 100 per week and count down 1 per position with a minimum score of 1.

Thus, 1st place gets 100 points, 100th place gets 1 point, and everybody after that gets 1 point. This minimizes the impact of waning participation.

This can be tweaked. Max and min scores can be adjusted.

[sikraiken]

The player who wins week one is guaranteed to win because the other weeks are not worth enough points.

You seem to be making some huge assumption that no one else is getting points during week one.


It doesn't matter; either way it's done, since it's linear, it'll be the same.

Someone who gets 1st, 1st, 10th, 10th, 20th will get the same number of points as someone who gets 10th, 10th, 20th, 1st, 1st.

We will take last year for example.

Week participation went like this:

108
93
88
77
76

Player X gets

99 for 10th
84 for 10th
69 for 20th
77 for 1st
76 for 1st

=405

Player Y gets
108 1st
93 1st
79 10th
68 10th
57 20th

=405


Let's do it 150 max.

Player X gets

141 10th
141 10th
131 20th
150 1st
150 1st

=713

Player Y gets

150 1st
150 1st
141 10th
141 10th
131 20th

=713

[DMC]

Now, I understand the problem with all weeks being equal. If each week is worth 150 points, and weeks one and two have 150 and 100 people respectively, that means you get 1 point for last place on week 1 and 51 points for last place on week 2.

Seems sad that you could get 50 points for submitting a terrible score just because nobody else played.

The fix: Low places are worth 1 each.

For example, start the score at 100 per week and count down 1 per position with a minimum score of 1.

Thus, 1st place gets 100 points, 100th place gets 1 point, and everybody after that gets 1 point. This minimizes the impact of waning participation.

This can be tweaked. Max and min scores can be adjusted

That sound like a really good idea worth consideration.

all though you could argue that if last place get an easy 51 points as you said in your example, that would my make people want to submit a score after all - thus participate... and in the end maybe that place won't be worth 51, but less.

So if you only get 1 point no matter what score it is (as long as youre not top-100), maybe you think it doesn't matter - and consequently you don't submit it?

But I really like this idea - after all I don't think so many more than will participate it's good to have this buffert if it happens to be 152 people participating in week 1 and less than 100 in some of the other weeks.

Edit: I would just like to add one thing:
sikraiken's example shows that this linear system is really good for top players. It doesn't matter if you place 1st in the first or last week, nor 10th or 20th. But for lower-tier players the linear system could be broken as discussed above. If you place 100ed it may matter if it's in the first or last week. But this could probably be fixed easily, maybe I give an example later.

[MR_Soren]

The player who wins week one is guaranteed to win because the other weeks are not worth enough points.

You seem to be making some huge assumption that no one else is getting points during week one.

Actually, I'm writing quickly at work and not thinking clearly.

It doesn't matter; either way it's done, since it's linear, it'll be the same.

Yes, if two players get the exact same results in a different order. But that won't usually be the case.

Back to your example:

108
93
88
77
76

New players:

Player A
2nd: 107
2nd: 92
2nd: 87
2nd: 76
Skip: 0

Total: 362


Player B
83rd: 26
1st: 93
1st: 88
1st: 77
1st: 76

Total: 360

Here, a player that wins four weeks and has one bad week is beaten by a player that takes second four times and completely skips the fifth week.

Yes, this example is contrived, but this illustrates the need for equal points every week.

[MX7]

150.000 GET

[sikraiken]

I should have added a little more, sorry.

So again, last year's numbers:

108
93
88
77
76

Player X gets

1 for 108th (Last)
1 for 93rd (Last)
5 for 84th (5th from last)
5 for 73rd (5th from last)
10 for 67th (10th from last)

=22

Player Y gets

5 for 104th (5th from last)
5 for 89th (5th from last)
10 for 79th (10th from last)
1 for 77th (Last)
1 for 76th (Last)

=22

Obviously if you want to compare each player place-for-place, you can't, because there is nothing below 93rd except in the first week, or below 88th besides the first and second week, etc. You can look at them as "equivalent" places since they each have two lasts, two 5th from lasts, and a 10th from last.


Let's do max 150:

Player X gets

108 for 108th (Last)
93 for 93rd (Last)
84 for 84th (5th from last)
73 for 73rd (5th from last)
67 for 67th (10th from last)

=425

Player Y gets

104 for 104th (5th from last)
89 for 89th (5th from last)
79 for 79th (10th from last)
77 for 77th (Last)
76 for 76th (Last)

=425

These situations all work for your typical player. If you have some strange/unlikely scenario like Mr_Soren pointed out, then yes, you'll have a problem. This happens because of nonexistant places in other weeks (there is no 83rd place in week 5). In that case, scoring based off an arbitrary max will be fairer. In majority of cases it would have been fair anyway (you more than likely aren't going to have a top player completely miss one week, or someone do terribly one week and then absolutely amazing all the other weeks). On top of that, if you care about rank, you'll participate each week. Max 150 (or whatever arbitrary number you use) is slightly better. Actually, going on player participation has no advantage over it that I can see.

[MR_Soren]

Max 150 (or whatever arbitrary number you use) is slightly better. Actually, going on player participation has no advantage over it that I can see.

Completely agree.

[Davey]

Are we officially going with a max score of 150, then?

[Enhasa]

Seems like there's some confusion on this page, but in terms of decision making there isn't much to discuss since every single person has indicated they want scoring by registrants (i.e. the top person gets the same score every week) instead of participants.

One point for each person you beat is honestly the best and simplest way to do it. (I made it clear in the example, but maybe I should have explicitly said it: non-participants count as a score of 0.) We have 152 registrants, which means that top scorer every week will be 151 (barring an extremely unlikely tie at the top spot). People who don't submit a score will get 0 points, as they will tie other non-players, but beat nobody.

[spongdangly]

Voting privately

This is an obvious yes. Voting through email where people put the game they want in the subject line would work. Apparently the PM box only holds 100 PMs, and we're past 100 participants.

Might I suggest that each team member pm's their votes to their team captain and then the captain pm's you (or whoever is in charge) the votes they received. This way you end up with just over 20 pm's (or e-mails) instead of 150.

[sven666]

is DDP2 playable in MAME?

if so it should definetly be added to a week.. "not taito" maybe?

[The Coop]

Voting privately

This is an obvious yes. Voting through email where people put the game they want in the subject line would work. Apparently the PM box only holds 100 PMs, and we're past 100 participants.

Might I suggest that each team member pm's their votes to their team captain and then the captain pm's you (or whoever is in charge) the votes they received. This way you end up with just over 20 pm's (or e-mails) instead of 150.

Sounds like a good idea to me.

**checks PM box**

Hmmm.... guess I better clean this out.

[spongdangly]

is DDP2 playable in MAME?

if so it should definetly be added to a week.. "not taito" maybe?

MAME 0.127 just came out today and the game is still not playable. Your ship turns invisible and you can't see enemy bullets either. However they have made progress since the last time I tried to play it.

[Fu]

At the 1st page of the thread, the games we are going to vote these weeks are definitve ?

[sikraiken]

Sorry I never answered your question Fu, I just saw it for the first time now.

Caravan week is going to lose Uwabami Breakers and pick up something else (which will be determined soon..). The other weeks aren't going to change.

The order of the weeks is definite - Capcom, Caravan, Not Taito, Doujin, Famous Manic. This is also posted in the Registration thread.

Also, scoring is 152-1 (152 points for first place, 1 point for 152nd).

[spongdangly]

Caravan week is going to lose Uwabami Breakers and pick up something else (which will be determined soon..).

Do you mean that Doujin week will be losing Uwabami Breakers? If so, thank you. I hate this game.

[sikraiken]

Yeah, I meant Doujin week. I was still waking up when I posted that...

[mikwuyma]

Sorry I never answered your question Fu, I just saw it for the first time now.

Caravan week is going to lose Uwabami Breakers and pick up something else (which will be determined soon..). The other weeks aren't going to change.

Is there a reason for this? Barring some sort of scoring exploit, I don't see any reason to drop the game. It does have enemy chaining, but the game is also quite short.

At the very least you could have told me about this

EDIT: What are you planning on replacing Uwabami with?

[Enhasa]

parsec47 roll extreme -nowait

[MX7]

parsec47 roll extreme -nowait