On Sun, 24 Apr 2022 09:22:33 +0100, David North
<nospam@lane-farm.fsnet.co.uk> wrote:

>Still a full round-robin to start with. The more rounds in the knockout
>stage, the less likely it is that the best team will win.

Depending on how you judge "best". I agree with what I take to be your
definition, so I'm not disagreeing here, but isn't it a circular
definition? If you define "best" as "the team which gets the most
points in the all-play-all stage", then obviously the best team will
be at the top of the league.

However, if the competition is defined to have a 9-round qualifying
stage to determine semi-finalists, why is it not a valid strategy to
make sure of qualification by winning 6 in a row, rest a load of
players for the last couple of rounds and drop to fourth place, and
then have a fully-rested and prepared team for the business end? And
if this strategy works, who is to say that they aren't the best team
even though they didn't finish top of the league?

I've already made clear that I don't approve of the grand final idea
for an all-play-all f-c competition, but I'm not sure that all the
arguments I'd feel like marshalling in favour of that position are
entirely sound.

Cheers,

Mike

On Sunday, April 24, 2022 at 8:13:54 PM UTC+10, Mike Holmans wrote:
> However, if the competition is defined to have a 9-round qualifying
> stage to determine semi-finalists, why is it not a valid strategy to
> make sure of qualification by winning 6 in a row, rest a load of
> players for the last couple of rounds and drop to fourth place, and
> then have a fully-rested and prepared team for the business end? And
> if this strategy works, who is to say that they aren't the best team
> even though they didn't finish top of the league?

If this strategy works (ie they win the final)?
Then the "best" is determined by the final....
Anyone saying "one of the losing team is best because they topped the table after the 9 qualifying rounds" would probably be named Robert.

If the team employing this strategy (win 6, rest squad, don't finish top) but DON'T win the finals.. would get vilified for resting their team rather than not securing a better semi-final placement. There would be endless comments about losing momentum etc.

On Monday, April 25, 2022 at 2:51:16 AM UTC+1, jzfre...@gmail.com wrote:
> On Sunday, April 24, 2022 at 8:13:54 PM UTC+10, Mike Holmans wrote:
> > However, if the competition is defined to have a 9-round qualifying
> > stage to determine semi-finalists, why is it not a valid strategy to
> > make sure of qualification by winning 6 in a row, rest a load of
> > players for the last couple of rounds and drop to fourth place, and
> > then have a fully-rested and prepared team for the business end? And
> > if this strategy works, who is to say that they aren't the best team
> > even though they didn't finish top of the league?
> If this strategy works (ie they win the final)?
> Then the "best" is determined by the final....
> Anyone saying "one of the losing team is best because they topped the table after the 9 qualifying rounds" would probably be named Robert.
>
> If the team employing this strategy (win 6, rest squad, don't finish top) but DON'T win the finals.. would get vilified for resting their team rather than not securing a better semi-final placement. There would be endless comments about losing momentum etc.

You either have a knockout competition or a league. Anything else is bogus competition. RH

On Monday, April 25, 2022 at 4:27:16 PM UTC+10, RH wrote:
> You either have a knockout competition or a league. Anything else is bogus competition. RH

That's quite close minded. Even for you.

Having said that, knockout is my least favourite. One bad day and you're gone.

I think RR with finals is a nice balance.
EPL works as it's so many games (20 teams, 2 RR).

Another format I have a lot of time for is double elimination. Basically you have to lose twice to get knocked out, with winners and losers brackets. They use it a lot in e-sports, and it works really well.
I don't watch much real sport these days, so I'm not sure if anyone is using the double elim.

On 2022-04-25, jack fredricks <jzfredricks@gmail.com> wrote:
> On Monday, April 25, 2022 at 4:27:16 PM UTC+10, RH wrote:
>> You either have a knockout competition or a league. Anything else is bogus competition. RH
>
> That's quite close minded. Even for you.
>
> Having said that, knockout is my least favourite. One bad day and you're gone.
>
> I think RR with finals is a nice balance.
> EPL works as it's so many games (20 teams, 2 RR).
>
> Another format I have a lot of time for is double elimination. Basically you have to lose twice to get knocked out, with winners and losers brackets. They use it a lot in e-sports, and it works really well.
> I don't watch much real sport these days, so I'm not sure if anyone is using the double elim.

Double elimination is used in the Schapiro Spring Fours, one of the
highlights of the English bridge calendar, which starts on Friday.

It's always puzzled me why this format isn't more widely used as
it is really not difficult to understand or implement.

I suspect that there aren't too many games of bridge played in the
pavilion by players on rainy days nowadays. Middlesex players had a
regular game in the days of Gatting and Edmonds.

On Sunday, 24 April 2022 at 11:13:54 UTC+1, Mike Holmans wrote:
> On Sun, 24 Apr 2022 09:22:33 +0100, David North
> <nos...@lane-farm.fsnet.co.uk> wrote:
>
>
> >Still a full round-robin to start with. The more rounds in the knockout
> >stage, the less likely it is that the best team will win.
> Depending on how you judge "best". I agree with what I take to be your
> definition, so I'm not disagreeing here, but isn't it a circular
> definition? If you define "best" as "the team which gets the most
> points in the all-play-all stage", then obviously the best team will
> be at the top of the league.

That wouldn't be my definition, and I'm even not sure that I would have a definition as such that could be determined from the results with certainty.

If you consider a theoretical best team, which has a higher probability of winning than losing against any other team, then it is more likely that that team will come to the top in a round-robin than in a one-off match or a series of knockout matches. Therefore the team that finishes at the top of the table is *more likely* to be the best team than the one who wins after semi-finals and a final (I did say 'probably').

> However, if the competition is defined to have a 9-round qualifying
> stage to determine semi-finalists, why is it not a valid strategy to
> make sure of qualification by winning 6 in a row, rest a load of
> players for the last couple of rounds and drop to fourth place, and
> then have a fully-rested and prepared team for the business end?

It is, although the team that finishes 4th usually plays the team that finishes top, so that strategy theoretically results in a more difficult semi-final.

> And
> if this strategy works, who is to say that they aren't the best team
> even though they didn't finish top of the league?

Agreed. In practice, it would be fairly well known, at the time at least, if a team had done that, so it could be taken into account in making a judgment.

I would take into account the results of the semi-finals and final in any case. E.g. if the top 4 in the table were closely bunched, it might be reasonable to judge the winner of the final as the best team, whereas if the top team finished a long way clear and then lost the semi-final or final narrowly, I would probably still consider them the best team.

On 25/04/2022 09:01, jack fredricks wrote:
[...]
> Another format I have a lot of time for is double elimination.
> Basically you have to lose twice to get knocked out, with winners and
> losers brackets.

Chess tournaments are commonly organised as a "Swiss system".
Think of it as a partial round robin; each round, you are paired against
someone [or a team] with the same number of points, as far as possible.
So the leading players essentially play in what amounts to a RR of the
strongest opposition, and the weakest players in what amounts to a RR of
the weakest players/teams. Full details left to the imagination/Wiki.

Pro:
No-one gets sent home after round one -- important for amateurs
who don't want to travel half-way across the UK [or Europe or the
world] at great expense to play just one afternoon.
No problem if players drop out, eg through illness/injury.
Good for determining the best players, even if they have the
odd bad day.
Can cope with large numbers of players in relatively few rounds
[eg 100 players and 7 rounds].
Weak players sometimes get the chance to play really strong
opposition [cf the FA Cup, where a lot of the interest comes from
seeing minnows playing, and occasionally beating, top teams], but
mostly you're playing opposition of roughly your own standard.
No need for preliminary rounds and similar -- subject to the
overall capacity of the venue, it doesn't matter how many players
decide to enter, nor how good/bad they are.

Con:
Many more matches than a KO [but fewer than a RR]. This is OK
for things like chess, less good for Wimbledon [where the 127
matches in (eg) the men's singles would be replaced by 7x64 ==
448 matches if the same players entered], where there is limited
capacity. Not a problem for cricket or football, where each team
has a venue anyway. [Wimbledon could cope with little change if
the main tournaments consisted of the 32 seeds plus 4 qualifiers,
for 7x18 == 126 matches. Juggle numbers to taste.]
You need to be able to organise fixtures at short notice.
Sports have to be able to do that anyway for KO, but this can
put pressure on venues.
For spectator sports, there are a lot of "dead" matches. OTOH,
this happens anyway in RR leagues, until you introduce artificial
extras such as "qualifying for Europe" or "play-off places".

It would be interesting to see the Football League run as a
Swiss. You could put the current FL teams plus the next few tiers
into a tournament with, say, 200 teams and 20 double [H&A] rounds.
The Liverpools and Man Cities would soon rise to the top, and would
play 40 matches against mostly the current Premiership, but every so
often a lesser team would have a good run and find themselves paired
against a top club, or a top team have a bad run and drop down to
mid-table.

--
Andy Walker, Nottingham.
Andy's music pages: www.cuboid.me.uk/andy/Music
Composer of the day: www.cuboid.me.uk/andy/Music/Composers/Ketterer

On Monday, April 25, 2022 at 11:38:11 PM UTC+10, Andy Walker wrote:
> No need for preliminary rounds and similar

read and snipped, thanks. An interesting system.

Prelim rounds seem needed for rankings though. With chess they can be played throughout the year ie prior to the tournament.
That's a tad trickier with cricket. So they'd probably just use rankings from the previous year.
One issue with that is it doesn't take into account squad changes (not an issue with 1 person chess, or even doubles tennis[where you can just average rankings]).

On 26/04/2022 00:18, jack fredricks wrote:
[Swiss system tournaments:]
> Prelim rounds seem needed for rankings though. With chess they can be
> played throughout the year ie prior to the tournament.
> That's a tad trickier with cricket. So they'd probably just use
> rankings from the previous year.
> One issue with that is it doesn't take into account squad changes
> (not an issue with 1 person chess, or even doubles tennis[where you
> can just average rankings]).

Just in case there's some confusion -- Swiss tournaments do not
normally need "rankings", in the sense of "top seed" and similar. Round
one can be typically paired at random; later rounds pair teams/players
with the same score /in the tournament/. So it doesn't matter how good
they are thought or measured to be before the event. Wiki describes a
number of variants, most of which do indeed use some form of seeding or
ranking, with the aim of speeding up the process whereby good teams
drift to the top of the table and weak teams to the bottom. But this
is important only if the number of rounds is no or little more than
the number needed for a KO [eg, 7 for 128 entrants], so is unlikely to
apply to any professional football or cricket league. Of course, if
the entry is restricted [eg because of the size/number of venues],
/then/ you need criteria for determining who is allowed to play; but
that's not really to do with the Swiss system, and already happens
with [eg] professional golf, tennis, snooker, ... tournaments.

[I found the Wiki article somewhat misleading, in that most of
it is about the variants rather than the basic system, and so gives an
undeserved emphasis to the importance of rankings.]

--
Andy Walker, Nottingham.
Andy's music pages: www.cuboid.me.uk/andy/Music
Composer of the day: www.cuboid.me.uk/andy/Music/Composers/Schubert