Roster Builder
After being in the hopper for almost two years, the site’s most customizable dashboard is finally here! Now you can upgrade from trying out individual lines to building an entire lineup with our Roster Builder.
Standings Points
The main goal of this visualization is to estimate how many points in the standings the fabricated team would likely attain. This is done using our SPAR (Standings Points Above Replacement) model which attempts to capture player contributions and is displayed in most of our other tools. Since the calculation is above replacement, we need to establish a baseline for what we would expect from a replacement-level team. With our process, this equates to roughly 41 standings points based on our standings projections for the 2024-24 season. This value may fluctuate over time and will be updated accordingly in the future.
Next up is the main component that drives results: the individual players. This section is broken down into the three main positions of forwards, defensemen, & goalies. You can also see breakdowns for each line-by-line on the left side of the viz where the SPAR is summed up for every forward line or defensive pair, then multiplied by the line’s time-on-ice multiplier before having its final line output displayed in the teal box. These TOI weights are determined by league tendencies, with the forward split for lines 1 through 4 being 19-16-14-11, while the defensive pairs get allocated 23.5-20.0-16.5 minutes. As for the goalies, a split of 54-28 starts is used between the selected starter and backup.
Speaking of the left side SPAR, underneath the teal player impacts are each line’s chemistry boost. These are the same values from the Line Compatibility Tool and look to capture how well linemates are expected to play together given their stylistic attributes.
The final element in this section is injuries (and fatigue). Given a player’s track record in the past couple years, we project how many games a player is expected to play in a given season and fill in the missed games with a replacement-level (0 SPAR) option. For many often-healthy skaters, they project near 82 and don’t lose much value by being substituted out, but injury-prone ones such as Mark Stone are closer to the mid-60s and are expected to lose roughly 20-25% of their value over a full season.
When it comes to the net-minders, this materializes in a “Fatigue” calculation. Since it’s difficult to calculate games missed due to injury for goaltenders (vs just being the backup by choice), a fatigue multiplier is applied if the user chooses to hand them more games than they are projected to start. Using this and league-wide tendencies, we can approximate how much this will impact a goalie’s performance as they face a heavier workload than they are acclimated to, and allocate this loss of SPAR to the injury section.
Team Identity
All of the above aspects are summed to get the team’s total Standings Points which features in the top right, and is used to characterize their likely level. For example, a team below with points in the 60s will get assigned “Lottery Candidate”, whereas one in the 89-97 range is in the “Wildcard Bubble” bracket.
The team’s total cap hit is also tracked, with an indicator if it does or doesn’t surpass the current salary cap ceiling. Once again similarly to the Line Compatibility Tool, we have the combined badges. These sum together the levels over every player and every badge to get standout skills throughout the team.
Lastly here is the “best two-linemate chemistry” which is exactly how it sounds. This simply goes over every forward line and defensive pair on the roster to identify the pair of players who sport the strongest connection on the team based on their playstyles. It should be noted that the players must be on the same line to count as an eligible combination here.
Customization
This dashboard has even more customization options to build specific teams. On the salary cap side of things, there is an option to add more (or remove) cap space from your team’s total. Whether that’s to represent retained contracts, buy-outs, healthy scratches, or other dead cap, you can manually adjust it here.
Below the viz and next to the manual cap option is a way to change the aforementioned ice time weights for each line. For example, if you wanted to opt for a more balanced approach for your top two defensive pairings, favour your first forward line more than usual, or alleviate your starter’s workload by giving some of his games to the backup, it can all be done here. To activate the custom resource allocation, simply click the checkbox and enter the TOI and GS values you wish. As is indicated in the sheet, if the proportions don’t add up to 60 minutes for the forward & defence lines, or 82 games started for the goalies, then they will automatically be scaled based on the ratio of numbers entered (ex: 4-3-2-1 scales up to 24-18-12-6).
Finally is probably the dashboard’s most complex feature. This one is located on the right-side boundary of the visualization and allows you to create custom players to enter into your roster that are not already in the LB-Hockey dataset. The process is still fairly short, here’s what’s needed:
1. Position: select the custom player’s position and enter the following info in the appropriate table
2. Name: this will only accept names that don’t already exist for the chosen position
3. SPAR: estimated standings points contribution over a full season
4. Cap: can be written as a number in millions, or with an “M”, “m”, “K”, “k” suffix (ex: 6.5M, 925k, 3.275)
5. GP Proj: for injury/fatigue calculations, games played for skaters and games started (GS) for goalies
6. Comparable: for chemistry calculations, the closest stylistic comparable from within the dataset
Once everything has been added for this player, they will be added to the dataset and available in the lineup drop-downs!
Note: these custom players are automatically assigned the NHL logo as their team and a blank headshot should they appear in the best chemistry section.