6.00
"Dynamic Replay adjustment (Chuck E. Cheese only)"
Wow! So Stern added a new feature, just for CEC, that implements
dynamic replay adjustment? By this, I'm guessing the machine
auto-percentages replays and adjusts the score automatically, like a
WMS game, without needing the start button to be hit at power-up?
That's pretty cool. How can Stern do this, given WMS's patent? Does
CEC have some special kind of deal allowing them to do this? (This
would surprise me very much, because CEC doesn't seem to care very
much about pinball....)
As we pretty much all know, High Speed was the first game to have
dynamic replay adjustment, in 1985. This gives WMS a patent lifetime
until 2005 (assuming the new 20-year patent law). Or does the old
17-year law still apply? If under the old law, the patent should have
expired. It's now 2003, 18 years since HS, so we're free! Does this
mean that Stern can also finally add compensation for broken switches!
That would be wonderful, and go a long way towards encouraging
operators to buy their machines.
7.00
"replays can now award a token/coin"
Wow, the way I read into this comment, it seems that payout pinball is
back!
I'm surprised people haven't made a bigger deal out of this.
Since most locations require multiple tokens to start a game, I like
the idea of giving the player back a token instead of just awarding a
credit on the machine. This allows a "partial Special" to be awarded:
instead of getting a completely free game, you merely get a discount
on your next game. Hmmm, deja vu for me:
http://groups.google.com/groups?selm=34c5be68.649221%40news.fix.net
But, since you get a physical token in your hand, you can apply your
winnings to another machine, if you're bored of playing the one you're
currently at! I like this idea very much.
If it's true that payout pinball machines are being built again, what
mechanism are they using to dispense tokens? It might be too
expensive, but it would be great if a coin mechanism can be licenced
from IGT, king of slot machines. They never jam! They have
experience designing a good coin mechanism that pays out as well as
takes in coins.
Discussion appreciated :)
Krellan
ARTMAN
>6.00
>"Dynamic Replay adjustment (Chuck E. Cheese only)"
>
>Wow! So Stern added a new feature, just for CEC, that implements
>dynamic replay adjustment? By this, I'm guessing the machine
>auto-percentages replays and adjusts the score automatically, like a
>WMS game, without needing the start button to be hit at power-up?
Actually, dynamic replay is completely different from anything that
games have done before (as far as I'm aware).
Before I go into dynamic replay, let me point out that our games
already do auto-adjust the replay since Harley Davidson, as we have
licensed the patent from wms. Start-button-pushing hasn't been
required in several years.
Our games have typically had "fixed", which means the game never
changes the score, and "auto", which is what I just described above
(which wms games have done for years).
The new "dynamic" replay system works thusly: You put in a base
replay, typically a fairly high number (say, 50M, which is high for
both RCT and SPP). You set the percentage of desired replay like
normal, 1-20% or whatever. In our example, we'll set it to 10%. In
order to achieve 10% replay, the game subtracts the desired percentage
of the base replay from the current replay amount. So, using the
above numbers, the replay starts at 50M when the game is first setup.
After one game is played, and the replay hasn't been won, the replay
becomes 45M. After the next game, it becomes 40M., and so on. The
replay will lower in this way until it is won or it hits the minimum
(typically 2M). When the replay is finally hit, the base replay is
added to the current replay. So if someone gets a replay at 10M
points, then on the next game, the replay is 60M and counts down from
there.
There are several advantages to this replay scheme. #1 is that it
spreads the replay out a LOT more evenly, so that even bad players get
a decent chance to hit it when it's pretty low. #2 is that it is MUCH
harder for a good player to absolutely kill the game because the
generally-bad clientelle has pulled the reflexing replay down to 3M or
whatever. It will actually make those special and HSTD credits more
valuable now. :)
There are probably disadvantages too, but the only major one I can
think of is this: When a replay is hit, the replay IMMEDIATELY goes
up, so if it is a multiplayer game, other players will have to work a
lot harder to get the replay! It was pretty much necessary to do it
this way, however, because increasing the score by 200M because all 4
players hit a 5M replay would suck a lot. But since the vast VAST
majority of games are 1-player games, it's really not all that big of
an issue.
This option is in fact available to the world (not just CEC) starting
on SPP, and is described in the manual. It is not on by default, an
operator will have to set their game up this way. However, I have set
2 of our test games up this way (including Season Tickets) and no one
seems to really loathe it. Or, at least, no one has slit my throat in
the parking lot yet because of it. I'm pretty happy with everything
it does, so there is an EXTREMELY good chance that by the next game I
do, it will be the default setting. I sincerely think every operator
should try it out, because I think it would help earnings
significantly.
>As we pretty much all know, High Speed was the first game to have
>dynamic replay adjustment, in 1985. This gives WMS a patent lifetime
>until 2005 (assuming the new 20-year patent law). Or does the old
>17-year law still apply? If under the old law, the patent should have
>expired. It's now 2003, 18 years since HS, so we're free! Does this
>mean that Stern can also finally add compensation for broken switches!
When the law was passed, patents got whatever ruling was most
favorable to them. Autopercentaging is around for at least another 3
years.
As for switch compensation, that infrastructure was written during the
development of HRC 3.00, and has been in every game since Playboy.
You can search google for the specifics (I talk about it a little bit
in the HRC 3.00 announcement), but basically we got a legal opinion
that said the method I came up with doesn't infringe the wms patent.
>If it's true that payout pinball machines are being built again, what
>mechanism are they using to dispense tokens? It might be too
>expensive, but it would be great if a coin mechanism can be licenced
>from IGT, king of slot machines. They never jam! They have
>experience designing a good coin mechanism that pays out as well as
>takes in coins.
It's my understanding that mostly the Germans are interested in this,
but who knows what some of the more clever/inventive ops will do with
the capability.
keith
--
We don't pay these people to serve as advertising agents for them.
Therefore, I suggest you do NOT use or support...
-----------== Posted via Newsfeed.Com - Uncensored Usenet News ==----------
http://www.newsfeed.com The #1 Newsgroup Service in the World!
-----= Over 100,000 Newsgroups - Unlimited Fast Downloads - 19 Servers =-----
*sigh* I wish the Chuck E. Cheese near me had a pinball machine. I'd be
a much happy camper when I take my kids there.
Now if only they served beer...
me: c'mon kids! Chuck E. Cheese again tonight!
kids: awwww, dad, we're _tired_ of Chuck E. Cheese!
A fellow can dream, can't he?
"Keith P. Johnson" <keith....@sternpinball.com> wrote in message
news:2u594v0c2va3j6nvj...@4ax.com...
>I have a minor gripe about the TOPS system, since we are talking about
>software. I think there should be an option to allow only one "win" per
>tournament. One I have has several players competing, but one guy always
>wins at least two of the top three places, pissing off the other players.
>Can an update be made to prevent this? It seems like hogging all the
>winnings. If you get a good score that places, but does not beat your own
>previous score, it would not qualify. If you beat it, it would replace it,
>and/or move you up the ranking. I just want it to be fair to all playing. We
>don't want to turn off players to the tournament.
Support for one win is in the works, but there's only so much we can
do. Ultimately it's up to the location to police their winners since
someone can easily put in different names, have their buddies collect,
etc.
keith
--
At one time I had random Your Mother jokes down here.
Maybe someday they'll return...
"Bryan Oakley" <br...@bitmover.com> wrote in message
news:3Ma1a.8$W91.10...@newssvr12.news.prodigy.com...
Damn.. you've just killed off one of the few benefits to playing location
games: piling up oodles of free credits thanks to pitifully low replay
reflexing! :) For a while, that became a game in itself.. find a machine
that works relatively well and with the replay bottomed out, and see how many
credits you can accumulate... Count both the total number of free games, and
the peak number of credits on the game during the run. :)
> There are probably disadvantages too, but the only major one I can
> think of is this: When a replay is hit, the replay IMMEDIATELY goes
> up, so if it is a multiplayer game, other players will have to work a
> lot harder to get the replay! It was pretty much necessary to do it
> this way, however, because increasing the score by 200M because all 4
> players hit a 5M replay would suck a lot. But since the vast VAST
> majority of games are 1-player games, it's really not all that big of
> an issue.
Could you have it not reflex til after the multiplayer game is over? Then
reflex it based on the best performing player of the four.
> As for switch compensation, that infrastructure was written during the
> development of HRC 3.00, and has been in every game since Playboy.
> You can search google for the specifics (I talk about it a little bit
> in the HRC 3.00 announcement), but basically we got a legal opinion
> that said the method I came up with doesn't infringe the wms patent.
Cool.. very cool... this'll help with those dead opto problems. Many a game
of HD was ruined by bad bike optos...
No, you wouldn't. The last one I was at had a pin in it. A Space Jam. One
flipper worked. Neither flipper had rubber on it. It was almost as much fun
as self immolation.
> Now if only they served beer...
This one did. It was like $8 for an 8 oz. plastic cup of cheap beer.
--
| David Gersic dgersic_@_niu.edu |
| Profanity is the linguistic crutch of inarticulate bastards |
| Email address is munged to avoid spammers. Remove the underscores. |
But what kind of _beer_ do they serve? -JW
Hate to burst your bubble but *I* really loathe it. I would have
complained to no end after playtesting at Season Tickets, but we
simply assumed this was a bug or some sort of braindead setting by the
operator.
It's really quite laughable to go up to a machine with a 2 million
replay score on it...barely clear that, then have the next game's
replay be set at a ridiculously unreachable amount. Percentages make
much more sense.
I think the Korn and i were joking about how silly a 50 million replay
was. We never imagined this was a "feature" and not a bug or some
hairbrained-operator's setting.
What it feels like to me, is cheap. It's almost like...oh. You got a
replay. Heh. Not again. Might as well just disable the replay
feature for a few games after a replay is awarded. Will have the same
effect to most players.
I agree that a great player abusing replays can be a problem, but I
really loathe this system Keith. I may be in the minority, but I just
wanted to let you know that it has been noticed.
> The new "dynamic" replay system works thusly: You put in a base
> replay, typically a fairly high number (say, 50M, which is high for
> both RCT and SPP). You set the percentage of desired replay like
> normal, 1-20% or whatever. In our example, we'll set it to 10%. In
> order to achieve 10% replay, the game subtracts the desired percentage
> of the base replay from the current replay amount. So, using the
> above numbers, the replay starts at 50M when the game is first setup.
> After one game is played, and the replay hasn't been won, the replay
> becomes 45M. After the next game, it becomes 40M., and so on. The
> replay will lower in this way until it is won or it hits the minimum
> (typically 2M). When the replay is finally hit, the base replay is
> added to the current replay. So if someone gets a replay at 10M
> points, then on the next game, the replay is 60M and counts down from
> there.
I'ld rather see something like where the ranges weren't so great. Using
your example, instead of dropping 5 million at a shot, it would drop 2
million... but when the game was won, it would only boost 20 million.
This would, IMO, seem less capricious, and a medium-skilled player
coming to play wouldn't have to wonder if he had a shot at winning-- he
would, because a few good games hadn't sent the replay level off into
the stratosphere. But it would still reflex fast enough to prevent
abuse from good players, and a run of bad players could still lower the
level down to where they had a shot at it.
Well, not _really_. I thought that the replay boost was simply set a
little haywire. You know, used to WMS style where if you hit a reply,
it goes up X million, or if it's below that, then they double the
replay score.
I didn't loathe it, but was simply amused by it. The speed at which
SPP reflexes the score is also amusing. (Have to give credit where
its due; I do think this will spread out replays far and wide, which
is a good thing.)
> What it feels like to me, is cheap. It's almost like...oh. You got a
> replay. Heh. Not again. Might as well just disable the replay
> feature for a few games after a replay is awarded. Will have the same
> effect to most players.
Yeah, but go play your T*Z and hit the replay. Now look at the replay
score. Same thing, different mechanism. (And yeah, different amount.
Big whoop.)
> I agree that a great player abusing replays can be a problem, but I
> really loathe this system Keith. I may be in the minority, but I just
> wanted to let you know that it has been noticed.
I dunno; I think it's actually better than the old WMS style
reflexing, which took WAY too long to reflex. I can go into just
about any arcade around here and stand on a machine for a good long
while on just a few, because everyone else has stunk up the house
soooo badly. This new hyper reflexing compensates a _lot_ for the
player that's sitting directly at the table, while still somewhat
taking into account how sucky (or dare I say good?) the "average"
customer is at that establishment. It's a good compromise in my book,
although amusing at times. :)
Back in the DE/Sega days we had an adjustment that could be used to
dispense a coin/token using the knocker position. IIRC the name of the
adjustment was "Italian Replay Knocker."
OD
--
Orin A. Day | "Hear the tick-tick, Bunny? *** How big is bunny?
lobster | So-o-o-o-o-o-o-o-o-o big! *** Bunny is eating his
@ | good supper. *** Sss-sh! Bunny is sleeping."
lobsterdevil.com | JUDY'S BOOK from Pat the Bunny by Dorothy Kunhardt
I thought that system worked really well and prevented excessive
replays without jumping the score by some nutty amount like 50
million.
the...@hotmail.com (The Korn) wrote in message news:<b2262ce8.03021...@posting.google.com>...
>I'ld rather see something like where the ranges weren't so great. Using
>your example, instead of dropping 5 million at a shot, it would drop 2
>million... but when the game was won, it would only boost 20 million.
>This would, IMO, seem less capricious, and a medium-skilled player
>coming to play wouldn't have to wonder if he had a shot at winning-- he
>would, because a few good games hadn't sent the replay level off into
>the stratosphere. But it would still reflex fast enough to prevent
>abuse from good players, and a run of bad players could still lower the
>level down to where they had a shot at it.
It's perfectly viable (and possible) to set your game up this way. I
simply chose 50M as the base amount, you can just as easily (actually
easier since you don't have to go as far :) ) set that number to be
20M. The trick is mainly finding a range that good players can't
seriously repeatedly win games. Whatever you set the base amount and
percentage amount to, the game will make it happen.