Deep dive: How does your side stack up in Rd 17?

• STAT ANALYSIS: Vital numbers to beat the beast
• CHAMPION PODS: Unique players to give you an edge

Analysing how the top 20,000 teams are positioned heading into round 17

For a while now, round 17 has loomed as a nightmare round. Bye rounds are always a challenge, but the lack of genuine keepers playing in round 17 has only been exacerbated by injuries over the past week or two.

So how well are you positioned relative to other sides? 7 players at this point probably sounds like a disaster. Just how bad is it and how does this vary based on how you are ranked? 

In this article I will seek to help you better understand how you stand relative to other sides. Hopefully it helps you in deciding whether you need to bring in two more guys playing this week or whether you can save a valuable trade for down the track.

PRICE DROP!!! Sign up to SC Playbook for just $20 for extra premium articles every round, including access to our subscriber special question and answer podcast.

Embed from Getty Images

Approach

To answer these questions, we will look into the detail of the top 20,000 teams – team-by-team.  

Following the announcement of team lists on Tuesday, we finally know who is playing this week – or at least we think we do. Teams are always subject to some late changes, and the one change I am assuming will happen is Joseph Suaalii coming in for Matt Ikuvalu in the much discussed jersey 21. So for the purpose of this article, I am assuming that change occurs.

The first step in our analysis is to then work out how many of those named players are in each of our 25 man squads. That’s pretty simple to do.  

The issue with that analysis is that it gives equal weight to a player such as Cody Walker as it does to a Ben Trbojevic – whereas in reality they should score very differently.

To adjust for this, I’ve given each player a projected score. I’ve typically used close to their season average but I’ve adjusted for where I think their opposition and/or position justifies it – for example, Joey Manu playing fullback against the Bulldogs. 

You may well have a different view of how players will score – I’ve just had a crack to do this analysis.

The tables below set out the scores I have assumed for the main SC relevant players in each match.

Results – Players Available in Round 17

The table below sets out the distribution of numbers of players available for selection in Round 17 by rank – prior to any trades this week. Note that no team in the top 20,000 has more than 16 players available!

In my case (the highlighted cell in the table below), I am just outside the top 1,000 and have 9 players available in round 17 pre-trades this week. 26% of teams ahead of me are in the same boat whereas 54% of teams have 8 or less players available.  

20% of teams in the top 1,000 have 10 or more players – but only 5% have 11 players and only 2% have 12 players. So I’m actually in reasonable shape despite only having 9 players at this point. 

In contrast, the 18,700 teams in the top 20,000 who are ranked behind me are less well placed – with only 12% of those teams having 10 or more players. So provided I get my captaincy selection right and my players score okay, I should see a ranking boost. More on that later.  

The final row in the table shows the average number of players available for selection by ranking zone. So, on average, the top 100 teams have 8.4 players available and the top 10,0000 teams have 7.5 – whereas the next 10,000 teams only have 6.8.  

The chart below sets out the average number of players available for the top 1,000 teams, the next 1,000 teams, the next 1,000 teams again and so on. As you can see, teams near the lead are much more prepared for the bye – but as the table shows, there are still plenty of teams in poor shape. Hence why bye rounds provide the greatest opportunity for ranking moves than any other rounds.

Results – Projected Scores in Round 17

So far so good. But how does the quality of my side stack up relative to other teams?

When I take my 9 players and the assumed scores in the tables earlier, I get to a projected score of 530 from my 9 players – prior to making any trades this week and excluding double points from my captain.

Applying that to every team in the top 20,000, we get the following average scores by ranking zone.

In other words, all other things being equal (i.e. captaincy choice and trades this week), the top 1,000 teams on average are projected to outscore the next 1,000 teams by 24 points…and those 1,000 teams are projected to outscore the teams ranked 2-3,000 by 19 points. So far my 530 points looks pretty good.

Whilst the averages are interesting, the distribution of scores is more useful. The graph below shows the distribution of scores amongst the top 20,000 teams grouped into 40 point ranges. So in my case, my 530 point projection means that I fall into the bottom quarter of the column labelled (520,560]. (Don’t blame me for the Excel formatting!).

That means that approx. 75% of that group of teams (i.e. 0.75 x 1,737) and all of the teams to the right on the graph (so 1,420 + 801 + 440…) have a higher projected score. So in total approx. 4,300 out of 20,000 teams are projected to score the same or more than me. I can tell you from the actual spreadsheet that I rank 4,215^th so close enough.

So to test your team, look up your available players in the tables above (pre trades) and then check your total against this distribution. Depending on your rank and your season objectives, anything towards the right of this distribution and you may be better to save a trade – especially those teams already forecasting 600+.

As one final piece, I then added up each team’s current score with their projected scores above to calculate their new total score to determine whether I think I’m expected to rise or fall down the rankings.  

In my case, I’m currently on 20,642 points so my projected score (excluding captain and trades) of 530 takes me to a new projected score of 21,172. When I do this across every team and rank them, my ranking increases from 1309^th to 1202^nd.   Pretty happy with that. On that basis I’ll probably make one trade and bank the other.

If you do the same, you can compare your new total against the table below to get a rough guide of your new rank as per the above. So if, for example, your new total using the approach above is 21,464 then that makes you about 400^th based on my projected scores. I’ve stopped at 10,000 because I only started with the top 20,000 teams and some of the teams outside of the top 20,000 will presumably move into the top 20,000.  

If you are really desperate to estimate your rank more accurately than that, I’ll use my score to give you a way to approximate it. In my case I’m projecting 21,172 which falls between 1,000 and 2,000^th in the table above.

The difference between a ranking of 1,000 and 2,000 is 232 points (i.e. 21,228 minus 20,996). My score of 21,172 is 176 points above the low end of that range (i.e. 21,172 minus 20,996).  

So I’m roughly 176/232 or 76% of the difference between the bottom and top of the rank.  

BREAK AWAY FROM THE PACK with Quantium, Australia’s premier data science and artificial intelligence company. Create your own career power play and join the team today. Visit our website for more information. CEO Adam Druissi is a contributor to SC Playbook.

So on that basis I would estimate my rank as 2000 – 760 = 1240. From the actual spreadsheet we already know that I rank 1202^nd so it’s pretty close. For what it’s worth, the difference between 1202^nd and 1240^th in rank is only 11 points.

Hopefully that helps you make decisions this week. If not, if you are still reading then hopefully you at least found it interesting! Best of luck..