A new competition format thread - you have been warned!

[PeterGustafsson]

Hi!


Reading through the recent NAC proposal thread finally got me to write down a competition format idea that I came up with quite some time ago.

I assert that the usual 3/4-poule to DE format is bad for team events, and the straight-to-DE format is even worse.

There are several reasons for this:
1. Those formats give little fencing to most contestants.
2. Apart from the teams in the very top, they do not give a good sorting.
3. Seeding of teams into poules of roughly equal strength ensures that there will be many blowouts in poules, even more so than in the 6-poules of individual fencing.
4. A single long match can hold up a lot of the rest of the schedule; as a consequence one often sees many empty pistes during team fencing, diminishing the usage efficiency.
5. They are dependent on reasonably good seeding information.
6. In the end, there will always be waiting time due to rest rules between semis and medal bouts.

Of course, these drawbacks are interrelated, and the relations between them have been discussed at length in previous threads.

So, here comes a proposal for a team format which does away with #3, and limits the scope of the other drawbacks except #6.

1. List the teams in a ordered list. The list can be ordered according to known seeding information, or it can be completely random. Both will work equally well.
2. All matches in round#1 follow one challenge format, while the matches in subsequent rounds follow another, simpler, format.
Format for round #1
3. The captain of the team which is lowest of the list challenges any team of his liking.
4. The captain of the challenged team can now either accept the challenge, or refuse it.
5. If the challenged captain accepts the challenge, those 2 teams go to the first available piste. The loser of that bout will be placed last in the standings after round #1, while the winner will be placed next to last. Go to #8.
6. If the challenged captain refuses the challenge, those 2 teams switch places in the standings. The refusing captain now becomes the captain of the team in the lowest spot on the list, and now becomes the challenger - Go to #4.
7. A challenged captain can not refuse more than one challenge since the last time two teams were sent to a piste. This rule is there to break otherwise endless cycles.
8. Once 2 teams have been sent to the 1st piste, the captain of the team that is lowest on the initial list of those teams that are not on the first piste becomes the challenging captain. Stages #3-7 go over again, until all teams (if the #teams is even) have been assigned opponents, or at most one team is left without opponent (if the #teams is odd). In the latter case, the team without opponent is placed on the top of the listings after round#1. Round #1 is now complete.
Format for subsequent rounds
9. The BC checks the bottom of the listings and strike off any teams that have amassed at least 3 losses. Those teams are done for the day, and keep the placement that they had at the round when they went out. The BC continues doing this until they find the lowest placed team that hass less than 3 losses. That team, and all teams placed above it, continue fencing at least one more round.
10. The captain of lowest surviving team issues a challenge to any team of his liking that his team has not fought before, but from now on the challenged team must take on the challenge. Those teams are sent to the first available piste.
11. The captain of the lowest ranked team except those that were sent to piste in stage #10 likewise challenges a team that he has not fought before, and those teams are sent to the next available piste.
12. Stage #11 is repeated until at most one surviving team has not been assigned an opponent. The BC is free to allocate pistes to whichever event they please throughout the competition. By moving pistes from one event to another, they can speed up whichever event that is going slowest.
13. Subsequent rounds are held until at most 4 teams are left surviving. At that team, the teams in spots #1-4 are placed in a semifinal bracket, which is fenced in its entirety nonwithstanding any previous matches between those teams.
14. STOP.

The big plus with this format is that nearly all bouts will be between reasonably evenly matched opponents. It works even without any formal seeding information, provided that the team captains have reasonably good hunches on relative team strengths. Since the BC does not have to do any strength comparison, no one can accuse them of creating suspiciously easy poules for one favored team. Whenever a piste becomes open, 2 teams are sent there, keeping piste occupancy high (The downside of that is that no team can go to a piste and know that they will be fencing just there for a long time.) The BC can allocate and reallocate pistes between events as a load balancing measure, and move things along accordingly - and/or see to it that all refs are being used in a weapon that they are rated in.

In order to limit the post size, I will provide an example in the next post.


Have a nice time!

Peter Gustafsson

[LordShout]

My initial thoughts on how to game the system, assuming random list, failed quite spectacularly because application of rule 7 and the place switching function will sort the list, assuming no team will fence a better team till they have to.

Clever Peter.

[PeterGustafsson]

Hi!


Imagine a competition run according to this format, in which the following teams take part: Jordan, France, Germany, Tunisia, Italy, Hungary.

First, the BC creates a random list. Let us assume that it comes out like this:

1. Jordan
2. Tunisia
3. Germany
4. Italy
5. Hungary
6. France

First thing: France gets to challenge, and goes for the easiest opponent, Jordan. The Jordanian captain has some self-preservation, and realizes that if he would accept the challenge he would in all likelyhood get himself an loss, and even if he would win, that would only place him #5 after round#1. So the Jordanian captain decines the challenge, and the two teams switch places on the list without having fenced. Both still have 0 losses, and the list now looks like this:

1. France
2. Tunisia
3. Germany
4. Italy
5. Hungary
6. Jordan

The Jordanian captain now gets to challenge, and he goes for Tunisia, which is his best bet for a win. The Tunisian captain realizes that this bout is also his best bet for a win, so he accepts the challenge. Both teams are sent off to fence on piste #1, and the list now looks like this:

1. France
2. Germany
3. Italy
4. Hungary

On piste#1: Tunisia - Jordan

Now, the Hungarian captain becomes the challenger, and he thinks that the strengths of his fencers are best suited to the Germans. However, the German captain thinks that the strengths of his fencers are best suited to the Italians, so he declines the challenge. The teams switch places without having fenced, and the list now looks like this:

1. France
2. Hungary
3. Italy
4. Germany

On piste#1: Tunisia - Jordan

The German captain becomes the challenger, and he obviously challenges the Italians. However, the Italian captain thinks that the strengths of his fencers are best suited to the Hungarians, so he declines the challenge. The teams switch places without having fenced, and the list now looks like this:

1. France
2. Hungary
3. Germany
4. Italy

On piste#1: Tunisia - Jordan

The Italian captain becomes the challenger, and he obviously challenges the Hungarians. Now, this could repeat forever unless there was a cycle-breaking mechanism, which is why I put in one. Hungary still wants to fence against Germany, so Italy and Hungary flip without having fenced. The list now looks like this:

1. France
2. Italy
3. Germany
4. Hungary

On piste#1: Tunisia - Jordan

Now, the Hungarian captain becomes the challenger, and he thinks that the strengths of his fencers are best suited to the Germans. He challenges the germans again, and since they already have been challenged before, the German captain can not decline the challenge. Both teams are sent off to fence on piste #2, and the list now looks like this:

1. France
2. Italy

On piste#1: Tunisia - Jordan
On piste#2: Germany - Hungary

The Italian captain becomes the challenger, and he obviously challenges the French, since there is no othe choice. The list now looks like this:

On piste#1: Tunisia - Jordan
On piste#2: Germany - Hungary
On piste#3: France - Italy


After round#1, France and Italy will have the top 2 spots, Germany and Hungary will have spots#3-4, while Tunisia and Jordan have the two last spots. Let us assume that Jordan, Germany, and Italy lose their respective bouts in the first round. Then the ranking after that looks like this:

1. France
2. Italy - 1 loss
3. Hungary
4. Germany - 1 loss
5. Tunisia
6. Jordan - 1 loss

Notice that we have arrived at a good ordering after just one round? No other format does such a quick ordering.

Bout assignments in the subsequent rounds go faster, since challenged captains can not decline bouts. The Jordanian captain again becomes the first challenger, and he picks Germany as the least bad alternative. They go off to piste #1. Then, the Tunisian captain challenges Hungary, and they go off to piste#2. Italy has already fenced against France, so piste #3 is free for any other concurrently running event. Assuming that the European teams win in round#2, the standings look like this after round#2:

1. France
2. Italy - 1 loss
3. Hungary
4. Germany - 1 loss
5. Tunisia - 1 loss
6. Jordan - 2 losses

The Jordanian captain yet again becomes the first challenger, and he picks Hungary as the least bad alternative. They go off to piste #1. Then, the Tunisian captain challenges Germany, and they go off to piste#2. Italy has already fenced against France, so piste #3 is free for any other concurrently running event. Assuming that the European teams win in round#3, the standings look like this after round#3:

1. France
2. Italy - 1 loss
3. Hungary
4. Germany - 1 loss
5. Tunisia - 2 losses
6. Jordan - 3 losses

Now, Jordan has amassed 3 losses, and is out. This leaves Tunisia as the first challenger, and its captain challenges Italy. Germany is the next challenger, and they can only challenge France. Assume that the challengers lose their bouts, then the list looks like this after round#4:

1. France
2. Italy - 1 loss
3. Hungary
4. Germany - 2 losses
5. Tunisia - 3 losses

Tunisia is out, and the 4 remaining teams are put into the semifinals, featuring France - Germany and Italy - Hungary.

This is the match count:
Round#1: 3
Round#2: 2
Round#3: 2
Round#4: 2
Total #matches before semis: 9

If this was fenced as a 6-poule, it would require 15 matches. If it were fenced as 2 3-poules, it would require 6 bouts, 4 of them blowouts, since a likely set of poules would be: (France, Germany, Tunisia) and (Italy, Hungary, Jordan)

If this 6-team event was fenced as a straight-to-DE format with additional DE tables for all teams eliminated in the same round. it would take 5 matches to get a winner, 1 more bout for the bronze medal, and 1 for the losers in the 1st round, giving a total of 7 matches. That is a little less than for my proposed format, but the nested-DE format is very prone to holdups - one long match delays things for the opponents of the winner of the long match, and the DE very rarely will schedule a match from another event in an open piste, since if that match is long, one will get avalanche effects.


Have a nice time!

Peter Gustafsson

[Bonehead]

...

It's like a modified insertion/bubble sort for fencers. I really like this idea. Once you have three losses your position can be assumed to be fixed.

It's a little bit complicated though, I suspect there is probably a way to game it, though I can't think of it.

[kalivor]

It's like a modified insertion/bubble sort for fencers. I really like this idea. Once you have three losses your position can be assumed to be fixed.

It's a little bit complicated though, I suspect there is probably a way to game it, though I can't think of it.

The gaming is to race to the bottom in the early rounds by fencing the bouts you don't mind losing, and then try to sneak into the seeding you want in the later rounds and the challenge options become fewer for opposing teams.

Here's an example that assumes that all the teams know their relative strengths, and try to win every bout. This results in a finals appearance between the two teams who best game the system:

1. Jordan
2. Tunisia
3. Germany
4. Italy
5. Hungary
6. France

Accepted as fact:
Tun > Jordan
EUR > Tun & Jordan
IT > FR
IT > DE
HU > IT
FR > HU
DE > FR
DE > HU

Instead of going with the easy wins early, however, everybody knows that they have losses to spare. They want to save their winnable bouts until later.

So France challenges not Jordan, but Germany. Who refuse.
1 - Jordan
2 - Tunisia
3 - France
4 - Italy
5 - Hungary
6 - Germany

Germany challenges Italy, who refuse.

1 - Jordan
2 - Tunisia
3 - France
4 - Germany
5 - Hungary
6 - Italy

Italy challenges Hungary, who refuse.

1 - Jordan
2 - Tunisia
3 - France
4 - Germany
5 - Italy
6 - Hungary

Hungary challenges Germany, who refuse.

1- Jordan
2 - Tunisia
3 - France
4 - Hungary
5 - Italy
6 - Germany

Germany challenges Italy again, who must accept. NOTE that everybody is trying to get to the bottom. The key to a good seeding is to be LOW in the rankings, and get to challenge in the late rounds ... without piling up losses.

Hungary challenges France, who refuse. France then challenges Tunisia, who accept. Everybody fences, with the results being:

1 - Hungary
2 - Jordan (1 loss)
3 - France
4 - Tunisia (1 loss)
5 - Italy
6 - Germany (1 loss)

Germany now gets to challenge first, with the certainty that they can beat anybody remaining, except Italy (who have trouble with Hungary). Germany challenges Hungary to both take the #1 seed and remove the ability for Italy to keep near the bottom of the rankings. Italy then challenges Tunisia, as they prefer to be #4 seed (if possible), thereby avoiding Hungary in the semis. France vs Jordan remains.

1 - Germany (1 loss)
2 - France
3 - Jordan (2 losses)
4 - Italy
5 - Tunisia (2 losses)
6 - Hungary (1 loss)

Hungary wants to avoid Germany until the finals, but can only beat Italy amongst the European countries. They figure that Tunisia MUST challenge Jordan. They take a risk, and challenge Italy for the #4 spot. Tunisia challenges Jordan (the only winnable bout available). France and Germany are the remaining bout.

1 - Germany (1 loss)
2 - France (1 loss)
3 - Tunisia (2 losses)
4 - Hungary (1 loss)
5 - Italy (1 loss)

Italy would like to challenge Tunisia. That would force Hungary to lose to France, and give them a chance to defeat France in the semi and Germany in the final. But they have already fenced. They are force to challenge France.

This, of course, is what Hungary was hoping for. They challenge Tunisia.

1 - Germany (1 loss)
2 - Italy (1 loss)
3 - Hungary (1 loss)
4 - France (2 losses)

The big winners were the ones who gamed the system the best: Germany and Hungary.

Hungary is a weak team, but positioned themselves to be against Italy (the only team they can beat) in the semi-finals.

Germany successfully positioned themselves to avoid facing the Italians (the only team they can't beat).

Italy is the big loser, as they are facing the only team present they cannot beat in the semi-finals. Their attempts to position themselves to face the Germans in the semi-finals have not just failed, but backfired.

[milstdfarm]

I'm no expert in team format but this seems really complicated.

What is the downside to a double elimination format with randomly assigned starting positions?

[Bonehead]

I'm no expert in team format but this seems really complicated.

What is the downside to a double elimination format with randomly assigned starting positions?

There are several reasons for this:
1. Those formats give little fencing to most contestants.
2. Apart from the teams in the very top, they do not give a good sorting.
3. Seeding of teams into poules of roughly equal strength ensures that there will be many blowouts in poules, even more so than in the 6-poules of individual fencing.
4. A single long match can hold up a lot of the rest of the schedule; as a consequence one often sees many empty pistes during team fencing, diminishing the usage efficiency.
5. They are dependent on reasonably good seeding information.
6. In the end, there will always be waiting time due to rest rules between semis and medal bouts.

As Peter said earlier.

[tchwojko]

Their attempts to position themselves to face the Germans in the semi-finals have not just failed, but backfired.

Having a system that requires good strategic choices in the system itself doesn't seem like a good system.

Why involve team choices at all? Why not force a set of challenges based on initial seeding?

[PeterGustafsson]

Hi!

A couple of points:

The gaming is to race to the bottom in the early rounds by fencing the bouts you don't mind losing, and then try to sneak into the seeding you want in the later rounds and the challenge options become fewer for opposing teams.

OK, now we have a spelled-out example on how to game the system.

Obvious drawback: The example contains only 6 teams and 36 possible matchups, making the strategy countable if one has good info both on relative team strengths, and the strategies used by opposing teams.

What if the competition is larger? Will it be possible to count through all possible sets of challenges and refusals, to arrive at the match order which is most beneficial to oneself? Brute force is probably not good enough - in my example, there were 9 matches before semifinals. If one assume that 6-team events give on average 9 matches before semis, that means that there will be at most 9!=362880 possible matchup orderings. With each extra competing team, the #possibilities grows extremely quickly.

What if one team uses a strategy, but fails to predict that another team uses a counterstrategy? What are the risks of a gaming strategy (as opposed to a naive strategy) backfiring? If one strategy seems efficient when it is used by one team against all other teams using a naive strategy (here defined are trying to get the easiest opposition avialable at any one time), would not the consequence be that all teams use that gaming strategy the next competition? If so, what then?

Here's an example that assumes that all the teams know their relative strengths, and try to win every bout. This results in a finals appearance between the two teams who best game the system:

1. Jordan
2. Tunisia
3. Germany
4. Italy
5. Hungary
6. France

Accepted as fact:
Tun > Jordan
EUR > Tun & Jordan
IT > FR
IT > DE
HU > IT
FR > HU
DE > FR
DE > HU

You have posited other strength relationships than the ones I used in post#2. Why? If you want to show a fundamentally different case, would it not be better to choose other country names, so as to prevent confusion?

Here are the relative strengths that I used in post#2:
FRA>(ITA,DE,HUN)>TUN>JOR
HU > DE
DE > ITA
ITA > HUN

Instead of going with the easy wins early, however, everybody knows that they have losses to spare. They want to save their winnable bouts until later.

I see two risks with this strategy. Firstly, a team using this strategy might find that the opponents that it sees as winnable matches have lost relatively early, leving them no good alternatives in the later rounds. Secondly, if one chooses an opponent which is tougher than the easiest alternative, one might get an unanticipated early loss. In a large competition with many rounds, that might well prove really costly.

So France challenges not Jordan, but Germany. Who refuse.
1 - Jordan
2 - Tunisia
3 - France
4 - Italy
5 - Hungary
6 - Germany

You posit that Germany is stronger than France. Would this not be a risky proposition on the french part, since they know that the germans could accept the challenge and then smack them? Where would that leave the French?

Germany challenges Italy, who refuse.

1 - Jordan
2 - Tunisia
3 - France
4 - Germany
5 - Hungary
6 - Italy

Italy challenges Hungary, who refuse.

1 - Jordan
2 - Tunisia
3 - France
4 - Germany
5 - Italy
6 - Hungary

Hungary challenges Germany, who refuse.

1- Jordan
2 - Tunisia
3 - France
4 - Hungary
5 - Italy
6 - Germany

Same comment against, thrice.

Germany challenges Italy again, who must accept. NOTE that everybody is trying to get to the bottom. The key to a good seeding is to be LOW in the rankings, and get to challenge in the late rounds ... without piling up losses.

But, if just one team breaks the pattern, will not the whole gaming break down?

Hungary challenges France, who refuse. France then challenges Tunisia, who accept. Everybody fences, with the results being:

1 - Hungary
2 - Jordan (1 loss)
3 - France
4 - Tunisia (1 loss)
5 - Italy
6 - Germany (1 loss)

Germany now gets to challenge first, with the certainty that they can beat anybody remaining, except Italy (who have trouble with Hungary). Germany challenges Hungary to both take the #1 seed and remove the ability for Italy to keep near the bottom of the rankings. Italy then challenges Tunisia, as they prefer to be #4 seed (if possible), thereby avoiding Hungary in the semis. France vs Jordan remains.

1 - Germany (1 loss)
2 - France
3 - Jordan (2 losses)
4 - Italy
5 - Tunisia (2 losses)
6 - Hungary (1 loss)

Hungary wants to avoid Germany until the finals, but can only beat Italy amongst the European countries. They figure that Tunisia MUST challenge Jordan. They take a risk, and challenge Italy for the #4 spot. Tunisia challenges Jordan (the only winnable bout available). France and Germany are the remaining bout.

1 - Germany (1 loss)
2 - France (1 loss)
3 - Tunisia (2 losses)
4 - Hungary (1 loss)
5 - Italy (1 loss)

Italy would like to challenge Tunisia. That would force Hungary to lose to France, and give them a chance to defeat France in the semi and Germany in the final. But they have already fenced. They are force to challenge France.

This, of course, is what Hungary was hoping for. They challenge Tunisia.

1 - Germany (1 loss)
2 - Italy (1 loss)
3 - Hungary (1 loss)
4 - France (2 losses)

The big winners were the ones who gamed the system the best: Germany and Hungary.

Hungary is a weak team, but positioned themselves to be against Italy (the only team they can beat) in the semi-finals.

Germany successfully positioned themselves to avoid facing the Italians (the only team they can't beat).

Italy is the big loser, as they are facing the only team present they cannot beat in the semi-finals. Their attempts to position themselves to face the Germans in the semi-finals have not just failed, but backfired.

Why did they not accept the 2nd challenge, when they got a chance to beat the Germans early on?


Have a nice time!

Peter Gustafsson

[PeterGustafsson]

Hi!

Having a system that requires good strategic choices in the system itself doesn't seem like a good system.

Why involve team choices at all? Why not force a set of challenges based on initial seeding?

1. If the BC can not control which matches will be fenced, nor their ordering, they can not be accused of giving unfair help to some team by creating an easy poule.
2. If there have been few team competitions involving the teams present, there will be little - or none at all - reliable quantifiable data on team matchups. However, individual fencers can be expected to assess their own team´s strength vs. at least some other opponents. Consider it crowdsourcing of seeding&matchup.
3. In some competitions, the relative strength differentials between each team and the best of those weaker than it might be fairly similar. In such evenly spaced competitions, the matchups involving adjacently ranked teams should be chosen, if one wants to maximize the proportion of bouts that are competitive, and not waste time on blowouts. In other competitions, the relative strength structure could be different - imagine a MET competition involving the national teams of France, Italy, Hungary, Germany, South Africa, Quatar, Jordan, and Iceland. In such a case one has two groups of teams, where the teams are roughly evenly matched inside a group, but all bouts between teams in different groups can be expected to be blowouts. In that case, it makes more sense to have as many matchups inside each group as possible. If one uses a predetermined (as I think you are implying) forced set of challenges, then such facts can not be accounted for.

That said, I have once run a small individual competition (7 competitors) in a forced-challenge format. It worked well, and I described both the format and the competition here on f.net.


Have a nice time!

Peter Gustafsson

[tchwojko]

1. If the BC can not control which matches will be fenced, nor their ordering, they can not be accused of giving unfair help to some team by creating an easy poule.

I was saying use your system, but the challenges are chosen algorithmically by initial seed. The only possible fudging by the Bout Committee is that seed, and even that ought to be done by objective criteria.

2. If there have been few team competitions involving the teams present, there will be little - or none at all - reliable quantifiable data on team matchups. However, individual fencers can be expected to assess their own team´s strength vs. at least some other opponents. Consider it crowdsourcing of seeding&matchup.

But it's not a very big crowd and requires playing a game before the game.

3. In some competitions, the relative strength differentials between each team and the best of those weaker than it might be fairly similar. In such evenly spaced competitions, the matchups involving adjacently ranked teams should be chosen, if one wants to maximize the proportion of bouts that are competitive, and not waste time on blowouts. In other competitions, the relative strength structure could be different - imagine a MET competition involving the national teams of France, Italy, Hungary, Germany, South Africa, Quatar, Jordan, and Iceland. In such a case one has two groups of teams, where the teams are roughly evenly matched inside a group, but all bouts between teams in different groups can be expected to be blowouts. In that case, it makes more sense to have as many matchups inside each group as possible. If one uses a predetermined (as I think you are implying) forced set of challenges, then such facts can not be accounted for.

That depends on whether you have a seeding based strictly on order or if you have something like a classification or rating system in place. (But that makes it more complicated.)

That said, I have once run a small individual competition (7 competitors) in a forced-challenge format. It worked well, and I described both the format and the competition here on f.net.

That doesn't mean that teams wouldn't start finding a way to game the system.

(I don't have any strong feelings about the format one way or another, just trying to find holes.)

[mrbiggs]

My main objection to this format (which, I admit, I haven't studied in enough detail to completely understand) is that it creates a game on top of fencing--in other words, you can be "good" at the competition format by choosing who to fence and who to deny appropriately. I find this to be a drawback over nearly all current formats, where who you fence is predetermined. If this is a modified insertion sort, I feel like it'd be possible to somehow automate the selections so that who fences who is correctly chosen automatically.

[kalivor]

You posit that Germany is stronger than France. Would this not be a risky proposition on the french part, since they know that the germans could accept the challenge and then smack them? Where would that leave the French?

Not at all!

If France fences Germany in the first round, there are two possible options:

1 - They BEAT Germany, and have won a bout they were not expecting to win.
2 - They lose to Germany (a loss they would suffer in any case), but remain as the bottom seed giving them the choice of opponents to challenge in round two. Additionally, if they move ahead of Germany in the rankings later, that German team will not be able to challenge them.

Note that, in my example, the two strongest teams are Germany and Italy. Germany positions themselves at the bottom after the first round, and finishes first overall in the final seedings.

Why did they not accept the 2nd challenge, when they got a chance to beat the Germans early on?

Sorry, I don't follow you there. Italy *does* beat Germany in the first round. They would probably have been better off if they had lost purposely. Then they could have challenged Hungary in the second round (losing again, staying at #6), and slowly moved up the rankings in an attempt to avoid Hungary in the semi-finals.

You are missing that there are two things of value in this pool:

1 - Winning, as this improves your seeding and keeps you from hitting the 3 loss knockout level.
2 - Losing, as this moves you towards the bottom of the rankings and improves the likelihood that you will be able to choose your next opponent.

In the early rounds, when teams are not near to elimination and have not fenced many other teams, the ability to choose future opponents is of more value than a good seeding. Which means that teams are not playing to WIN, but jockeying for position. If you are a front-runner (#1 or #2 seed position), there is no scenario in which you get to choose your opponent.

This is not unlike a medium- or long-distance race, in which runners do not want to be the guy in front, but merely within striking distance of that runner.

Winning your way to the front of the pack in the early rounds simply burns up wins while painting a target on your back.